INTERNATIONAL ORGANISATION FOR STANDARDISATIONORGANISATION INTERNATIONALE DE NORMALISATION

ISO/IEC JTC 1/SC 29/WG 11CODING OF MOVING PICTURES AND AUDIO

ISO/IEC JTC 1/SC 29/WG 11 N8913San José, CA, US – April 2007

Source: Leonardo Chiariglione Title: Report of 80th meetingStatus

Report of 80th meeting

1 OpeningThe 80th meeting was held at the invitation of ANSI and held at the San José Double Tree Hotel.

2 Roll call of participantsAnnex 1 gives the attendance list

3 Approval of agendaAnnex 2 gives the approved agenda

4 Allocation of contributionsAnnex 3 gives the list of input documents

5 Communications from ConvenorThere was no specific communication

6 Report of previous meetingThis was approved

7 Processing of NB Position PapersNB position papers were considered and responses provided where appropriate

8940 Response to National Bodies

8 Work plan

8.1 Media coding

8.1.1 MPEG-4 Visual Simple Profile Level 6The following documents were approved

8948 Disposition of Comments on ISO/IEC 14496-2:2004/PDAM4

1

8949 Text of ISO/IEC 14496-2:2004/FPDAM4 Simple Profile Level 6

8.1.2 Scalable Video CodingThe following documents were approved

8962 Study Text (version 3) of ISO/IEC 14496-10:2005/FPDAM3 Scalable Video Coding8963 Joint Scalable Video Model (JSVM) 108964 JSVM 10 Software8965 Draft SVC Verification Test Plan Version 3.0

8.1.3 Multiview Video CodingThe following documents were approved

8966 Working Draft 3 of ISO/IEC 14496-10:2005/Amd.4 Multiview Video Coding

8967 Joint Multiview Video Model (JMVM) 48968 JMVM 4 Software

8.1.4 AAC-ELDThe following documents were approved

9072 DoC on ISO/IEC 14496-3:2005/PDAM 9 Request for Amendment.9073 DoC on ISO/IEC 14496-3:2005/PDAM 99074 ISO/IEC 14496-3:2005/FPDAM 9, AAC-ELD

8.1.5 Geometry and Shadow The following documents were approved

9136 WD 2.0 of ISO/IEC 14496-16:2006/AMD2 (Frame-based Animated Mesh Compression)9150 Request for ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)9137 WD 1.0 of ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)9138 3D Graphics Core Experiments Description

8.1.6 Video Tool LibraryThe following document was approved

8984 WD 4 of ISO/IEC 23002-4

8.1.7 Bitstream Syntax Description LanguageThe following documents were approved

9127 Text of ISO/IEC 23001-5 FDIS Bitstream Syntax Description Language

8.1.8 Fixed point implementation of DCT/IDCTThe following documents were approved

8982 Disposition of Comments on ISO/IEC CD 23002-28983 Text of ISO/IEC FCD 23002-2 Fixed-point 8x8 IDCT and

DCT

2

8.1.9 Spatial Audio Object Coding The following documents were approved

9099 Final Spatial Audio Object Coding Evaluation Procedures and Criterion9090 DoC ISO/IEC 23003-1:2007/PDAM 1

8.1.10 Free Viewpoint TV codingThe following documents were approved

8944 FTV Model and Requirements

8.1.11 Audio and speech codingThe following documents were approved

9095 Framework for Exploration of Speech and Audio Coding9096 Workplan for Exploration of Speech and Audio Coding

8.2 Composition coding

8.2.1 Lightweight Scene RepresentationThe following documents were approved

9028 DoC on ISO/IEC 14496-20/FPDAM1 (LASeR Extensions)9029 Text of ISO/IEC 14496-20/FDAM1 (LASeR Extensions)9030 Request for ISO/IEC 14496-20/Amd.2 (SVGT1.2 Support)9031 Text of ISO/IEC 14496-20/FPDAM2 (SVGT1.2 Support)9032 TuC for ISO/IEC 14496-20/Amd29033 WD3.0 of ISO/IEC 14496-20 2nd Edition (1st Ed. + Cor + Amd.1)9034 Ideas under Consideration (IuC) for LASeR

8.2.2 Symbolic Music RepresentationThe following documents were approved

9088 DoC ISO/IEC FCD 14496-239089 ISO/IEC FDIS 14496-23:200x, Symbolic Music Representation

8.3 Description Coding

8.3.1 Schema definitionThe following documents were approved

9102 Schema Files for MPEG-7

8.3.2 Visual Descriptor ExtensionsThe following documents were approved

8970 MPEG-7 Visual XM Document version 30.08971 Description of Core Experiments for MPEG-7 New Visual

Extensions

3

8.3.3 Improvements to Geographic DescriptorThe following documents were approved

9129 DoC on ISO/IEC PDAM/3 15938-5 Improvements to Geographic Descriptor

9100 ISO/IEC FPDAM/3 15938-5 Improvements to Geographic Descriptor

8.3.4 MPEG-7 Query FormatThe following documents were approved

9151 Request for subdivision ISO/IEC 15938-12 MPEG-7 Query Format9103 ISO/IEC 15938-12 CD MPEG-7 Query Format9104 Technologies Under Consideration for MPEG-7 Query Format

8.4 Systems support

8.4.1 Fragments Request UnitThe following documents were approved

9050 DoC on ISO/IEC 23001-2/FCD (Fragment Request Unit)9051 Text of ISO/IEC 23001-2/FDIS (Fragment Request Unit)

8.5 IPMP

8.5.1 IPMP XML Messages The following documents were approved

9052 Text of ISO/IEC 23001-3/FCD (IPMP XML Messages)9144 TuC for IPMP XML Messages

8.5.2 MPEG-21 IPMP Component Base ProfileThe following documents were approved

9105 DoC of ISO/IEC 21000-4 FPDAM/1 IPMP Components Base Profile9106 Text of ISO/IEC 21000-4 FDAM/1 IPMP Components Base Profile

8.5.3 REL Open Release ProfileThe following documents were approved

9107 DoC of ISO/IEC 21000-5 PDAM/3 ORC (Open Release Content) Profile9108 ISO/IEC 21000-5 FPDAM/3 ORC (Open Release Content) Profile

8.5.4 REL Distribution and Capture ProfileThe following document was approved

9109 Interoperability between MPEG-21 REL DAC Profile and other Rights Information Standards

4

8.6 Digital Item

8.6.1 Digital Item AdaptationThe following document was approved

9113 Text of ISO/IEC 21000-7 FDIS Second edition

8.7 Transport and File Format

8.7.1 Transport of MPEG Surround data in AACThe following documents were approved

9066 DoC ISO/IEC 13818-7:2006/FPDAM 19067 ISO/IEC 13818-7:2006/FDAM 1, Transport of MPEG Surround data in AAC

8.7.2 Flute Hint TrackThe following documents were approved

9022 DoC on ISO/IEC 14496-12/FPDAM2 (Flute Hint Track)9023 Text of ISO/IEC 14496-12/FDAM2 (Flute Hint Track)9025 TuC for ISO/IEC 14496-12 & 15444-12

8.7.3 AVC File Format extensions for SVCThe following documents were approved

9026 Study Text of ISO/IEC 14496-15/PDAM2 (SVC File Format)

8.7.4 MP4FF box for Original Audio File InformationThe following documents were approved

9070 DoC on ISO/IEC 14496-3/PDAM 89071 ISO/IEC 14496-3/FPDAM 8, MP4FF Box for Original Audio File

Information

8.7.5 Digital Item File FormatThe following documents were approved

9035 Request of ISO/IEC 21000-9/Amd.19036 Text of ISO/IEC 21000-9/PDAM.1 Mime Type Registration

8.7.6 Digital Item StreamingThe following documents were approved

9119 DoC of ISO/IEC 21000-18/PDAM 19120 ISO/IEC 21000-18/FPDAM/1 Simple fragmentation rule

5

8.8 Multimedia architecture

8.8.1 M3W Component DownloadThe following document was approved

9053 Text of ISO/IEC 23004-5/FDIS Component Download

8.8.2 M3W Fault ManagementThe following document was approved

9054 Text of ISO/IEC 23004-6/FDIS Fault Management

8.8.3 M3W System Integrity ManagementThe following document was approved

9055 Text of ISO/IEC 23004-7/FDIS System Integrity Management

8.8.3.1 Codec Configuration RepresentationThe following documents were approved

8979 WD 4 of ISO/IEC 23001-48985 Description of Core Experiments in RVC8986 RVC Simulation Model (RSM) V4.08987 RVC Work Plan8989 Description of Exploration Experiments for Toolbox Extensions

8.8.4 3D Graphics Compression ModelsThe following documents were approved

9141 Request for Subdivision of ISO/IEC 14496: Part 25 - 3D Graphics Compression Model9142 WD 1.0 for ISO/IEC 14496-25

8.8.5 Media Streaming MAF ProtocolsThe following documents were approved

9058 DoC on ISO/IEC 29116-1/CD Media Streaming MAF Protocol9059 Text of ISO/IEC 29116-1/FCD Media Streaming MAF Protocol

8.8.6 Extensible Multimedia PlatformThe following documents were approved

9060 A project to exploit MPEG standards in tune with industry practices and needs

8.9 Application formats

8.9.1 Protected Music Player MAFThe following documents were approved

9121 DoC of ISO/IEC 23000-2 FCD Music Player Application Format 2nd Edition9122 Text of ISO/IEC 23000-2 FDIS Music Player Application Format 2nd Edition

6

8.9.2 Musical Slide Show MAFThe following documents were approved

9037 DoC of ISO/IEC FCD 23000-4 (Musical Slide Show MAF)9038 Text of ISO/IEC FDIS 23000-4 (Musical Slide Show MAF)9040 WD1.0 of ISO/IEC 23000-4/Amd.2 Protected Musical Slide Show

8.9.3 Media Streaming MAFThe following documents were approved

9123 DoC on ISO/IEC CD 23000-5 Media Streaming Player9124 ISO/IEC FCD 23000-5 Media Streaming Player

8.9.4 Open Release Application FormatThe following documents were approved

9125 DoC of ISO/IEC 23000-7 CD Open release MAF9126 ISO/IEC 23000-7 FCD Open release MAF

8.9.5 Portable Video PlayerThe following documents were approved

9041 Text of ISO/IEC 23000-8/CD (Portable Video Player MAF)

8.9.6 Digital Multimedia Broadcasting Application FormatThe following documents were approved

9042 DoC on ISO/IEC 23000-9/CD (MAF for DMB)9043 Text of ISO/IEC 23000-9/FCD (MAF for

DMB)9044 TuC on MAF for DMB

8.9.7 Video Surveillance Application FormatThe following documents were approved

9045 Request for ISO/IEC 23000-109046 WD1.0 on ISO/IEC 23000-10 (Video Surveillance MAF)

8.10 Reference implementation

8.10.1 File Format Reference SoftwareThe following documents were approved

9019 DoC of ISO/IEC 14496-5/FPDAM12 File Format Reference Soft.9020 Text of ISO/IEC 14496-5/FDAM12 File Format Reference Software

8.10.2 Reference Hardware DescriptionThe following documents were approved

8994 Status of HDL submissions and commitments for MPEG

7

8995 Study of ISO/IEC DTR 14496-9

8.10.3 Geometry and Shadow Reference SoftwareThe following documents were approved

9149 Doc of ISO/IEC 14496-5:2001/ PDAM13 (Geometry and Shadow RefSoft)9135 Text of ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)

8.10.4 MPEG-J GFX Reference SoftwareThe following documents were approved

9148 Doc of ISO/IEC 14496-5:2001/ FPDAM11 (MPEG-J GFX RefSoft)9134 Text of ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)

8.10.5 New Profiles for Professional Applications Reference SoftwareThe following documents were approved

8958 Request for ISO/IEC 14496-5:2001/Amd.188959 Working Draft 1 of ISO/IEC 14496-5:2001/Amd.18 Reference Software for new Profiles for

Professional Applications

8.10.6 SVC Reference SoftwareThe following documents were approved

8960 Request for ISO/IEC 14496-5:2001/Amd.198961 Working Draft 1 of ISO/IEC 14496-4:2001/Amd.19 Reference Software for

SVC

8.10.7 BSAC Reference SoftwareThe following documents were approved

9086 Request for Amendment, MPEG-1/2 on MPEG-4 Ref. Software

9087 Text of ISO/IEC 14496-5:2001/PDAM 20, MPEG-1/2 on MPEG-4 Ref. Software

8.10.8 Perceptual 3D Shape Reference SoftwareThe following documents were approved

8974 Disposition of Comments on ISO/IEC 15938-6:2003/FPDAM28975 Text of ISO/IEC 15938-6:2003/FDAM2 (Perceptual 3D Shape)

8.10.9 Rights Expression Language Reference SoftwareThe following documents were approved

9110 REL/RDD Reference Software Development Plan v.6

8.10.10 Digital Item Reference Software The following documents were approved

9114 Preliminary DoC of preliminary comments of ISO/IEC 21000-8 FCD Reference Software

8

9115 Study text of ISO/IEC 21000-8 FCD Reference Software

8.10.11 Rights Data Dictionary Reference SoftwareThe following documents were approved

9110 REL/RDD Reference Software Development Plan v.6

8.10.12 Photo Player MAF Reference SoftwareThe following documents were approved

8978 Study Text of ISO/IEC 23000-3/PDAM1 Reference Software for Photo Player MAF

8.10.13 Musical Slide Show MAF Reference Software

9039 Workplan for Musical Slide Show MAF Conformance and Ref. Software

8.10.14 Prefixes and wild card extensions Reference SoftwareThe following documents were approved

9047 Study Text of ISO/IEC 23001-1/FPDAM2 (Prefixes and of wild cards extensions)

8.10.15 Integer IDCT Accuracy Testing Reference SoftwareThe following documents were approved

8980 Disposition of Comments on ISO/IEC 23002-1/PDAM18981 Text of ISO/IEC 23002-1/FPDAM1 Software for Integer IDCT Accuracy Testing

8.10.16 MPEG Surround Reference SoftwareThe following documents were approved

9093 ISO/IEC 23003-1:2007/FPDAM 2, MPEG Surround Reference Software9094 Defect Report of ISO/IEC 23003-1:2007

8.10.17 M3W Reference SoftwareThe following documents were approved

9056 WD2.0 of ISO/IEC 23004-8 Reference Software and Conformance9057 M3W Reference Software and Conformance Plan

8.11 Conformance

8.11.1 File Format ConformanceThe following documents were approved

9013 DoC on ISO/IEC 14496-4/PDAM 24 File Format Conformance9014 Text of ISO/IEC 14496-4/FPDAM 24 File Format Conformance

8.11.2 Geometry and Shadow ConformanceThe following documents were approved

9

9147 DoC of ISO/IEC 14496-4:2001/ PDAM21 (Geometry and Shadow Conformance)

9133 Text of ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

8.11.3 Synthesised Texture Conformance The following documents were approved

8999 DoC on ISO/IEC 14496-4/PDAM 23 Synthesised Texture Conformance9012 Text of ISO/IEC 14496-4/FPDAM 23 Synthesised Texture Conformance

8.11.4 MPEG-J GFX ConformanceThe following documents were approved

9146 DoC of ISO/IEC 14496-4:2001/ FPDAM16 (MPEG-J GFX Conformance)

9132 Text of ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

8.11.5 Laser ConformanceThe following documents were approved

9015 DoC on ISO/IEC 14496-4/PDAM 25 LASeR V1 Conformance9016 Text of ISO/IEC 14496-4/FPDAM 25 LASeR V1 Conformance

8.11.6 Open Font Format ConformanceThe following documents were approved

9017 Request for ISO/IEC 14496-4/Amd.269018 Text of ISO/IEC 14496-4/PDAM 26 Open Font Format Conformance

8.11.7 Visual Simple Profile Level 6 Conformance The following documents were approved

8952 Disposition of Comments on ISO/IEC 14496-4:2004/PDAM28 8953 Text of ISO/IEC 14496-4:2004/FPDAM28 Visual Simple Profile Level 6 Conformance

Testing

8.11.8 New Profiles for Professional Applications ConformanceThe following documents were approved

8954 Request for ISO/IEC 14496-4:2004/Amd.30 8955 Working Draft 1 of ISO/IEC 14496-4:2004/Amd.30 Conformance Testing for new

Profiles for Professional Applications

8.11.9 SVC Profiles ConformanceThe following documents were approved

8956 Request for ISO/IEC 14496-4:2004/Amd.31 8957 Working Draft 1 of ISO/IEC 14496-4:2004/Amd.31 Conformance Testing for SVC

Profiles

10

8.11.10 MPEG-1 and -2 Audio in MPEG-4 ConformanceThe following documents were approved

9078 DoC ISO/IEC 14496-4:2004/FPDAM 18

9079 ISO/IEC 14496-4:2004/FDAM 18, MPEG-1 and -2 on MPEG-4 Conformance

8.11.11 BSAC ConformanceThe following documents were approved

9076 DoC on ISO/IEC 14496-4:2004/FPDAM 149077 ISO/IEC 14496-4:2004/FDAM 14, BSAC Extensions Conformance

8.11.12 Audio Lossless ConformanceThe following documents were approved

9080 DoC ISO/IEC 14496-4:2004/FPDAM 199081 ISO/IEC 14496-4:2004/FDAM 19, ALS Conformance

8.11.13 Perceptual 3D Shape ConformanceThe following documents were approved

8976 Disposition of Comments on ISO/IEC 15938-7:2003/FPDAM38977 Text of ISO/IEC 15938-7:2003/FPDAM3 (Perceptual 3D Shape)

8.11.14 Improvements to Geographic Descriptor Conformance

9130 DoC on ISO/IEC PDAM/4 15938-7 Improvements to Geographic Descriptor Conformance9101 ISO/IEC FPDAM/4 15938-7 Improvements to Geographic Descriptor Conformance

8.11.15 Digital Item ConformanceThe following documents were approved

9116 DoC of ISO/IEC 21000-14 Conformance9117 Text of ISO/IEC FDIS 21000-14 Conformance

8.11.16 Musical Slide Show MAF Conformance The following document was approved

9039 Workplan for Musical Slide Show MAF Conformance and Ref. Software

8.11.17 MPEG Surround ConformanceThe following document was approved

9091 ISO/IEC 23003-1:2007/FPDAM 1, MPEG Surround Conformance9092 DoC ISO/IEC 23003-1:2007/PDAM 2

8.11.18 Codec Configuration Representation ConformanceThe following document was approved

11

8988 RVC Conformance Testing Working Draft 1.0

8.12 Maintenance

8.12.1 Systems coding standardsThe following documents were approved

8998 Text of ISO/IEC 13818-1:2003/DCOR1.2 (AVC Referencing and PS Signalling)9021 Text of ISO/IEC 14496-11/COR.6 (AudioFXProto correction and Bitwrapper)9024 Text of ISO/IEC 14496-12/COR.39027 ISO/IEC 14496-20/DCOR29140 Text of ISO/IEC 14496-21:2006/COR18972 Disposition of Comments on ISO/IEC 15938-6:2003/ Amd.1:2006/DCOR 18973 Text of ISO/IEC 15938-6:2003/Amd.1:2006/Cor.1 (Color Temperature)9048 DoC on ISO/IEC 23001/DCOR29049 Text of ISO/IEC 23001/COR2

8.12.2 Video coding standardsThe following documents were approved

9064 DoC on ISO/IEC 11172-5:199x/DCOR 19065 ISO/IEC 11172-5:199x/Cor. 18950 Text of ISO/IEC 14496-4:2004/DCOR4 8951 Text of ISO/IEC 14496-4:2004/Amd.1/DCOR2

8.12.3 Audio coding standardsThe following documents were approved

9068 ISO/IEC 14496-3:2005/DCOR 5 (DST and MP3on4)9069 ISO/IEC 14496-3:2005/DCOR 6 (SLS)9085 Text of ISO/IEC 14496-5:2001/AMD 10:2007/DCOR 1, BSAC and SLS

8.12.4 Visual description coding standardsThe following documents were approved

8969 Text of ISO/IEC 15938-3:2002/Amd.2:2006/Cor.1 (Perceptual 3D Shape)

8.12.5 Digital Item standards

9111 Disposition of Comments on ISO/IEC 21000-7:2004/DCOR 19118 ISO/IEC 21000-15:2006/DCOR1 MPEG-21 Event Reporting

9 Liaison mattersThe following output liaisons were issued

8919 Liaison statement to WG18920 Liaison Statement to IETF8921 Liaison Statement to Khronos

12

8922 Liaison Statement to ISO TC184 SC48923 Liaison Statement to 3GPP8924 Liaison Statement to W3C8925 Liaison Statement to ITU-T FG/IPTV concerning M3W8926 Liaison Statement to ITU-T FG IPTV8927 Liaison Statement to SMPTE 8928 Liaison Statement to DVD Forum8929 Liaison Statement to ETSI 8930 Liaison Statement to SMPTE re file format 8931 Liaison Statement to DVB 8932 Liaison Statement to JCP8933 Liaison Statement to CEA8934 Liaison Statement to ATIS8935 Liaison Statement to SMPTE re RVC8936 Liaison Statement to 3D Consortium8937 Liaison Statement to FLOForum8938 Liaison Statement to TC46/SC9/WG78939 Liaison Statement to AVS8941 Liaison Statement to DVB

10 Organisation of this meeting

10.1 Tasks for subgroupsThe following tasks were assigned to subgroups

S P AReqs

4 16 3D compression profiling20 Laser profiling

21 2 New DIDA 3 MAFs under consideration: Protected Photo Player

4 MAFs under consideration: Protected Musical Slide Show5 MAFs under consideration: Digital Cinema

10 MAFs under consideration: SurveillanceY Stereoscopic MAFZ Cross media interactive presentation

C 4 RVC Toolbox ExtensionMPEG URNsMAF Awareness EventFTV

Systems4 4 22 Audio BIFS conformance

23 Synthesised texture conformance24 File format conformance25 Laser conformance26 Open Font Format Conformance2x Laser v.2 conformance

5 12 File Format Reference Software14 Open Font Format Reference Software

13

16 Symbolic Music Representation Reference Software17 Laser Reference Software

12 2 FLUTE hint track15 1 SVC File Format20 1 Lightweight Scene Representation

21 9 1 Mime type registrationA 4 Musical Slide Show MAF

1 Musical Slide Show MAF conformance & RS2 Protected Musical Slide Show MAF

8 Portable Video Player MAF9 DMB MAF

10 Video Surveillance MAFB 1 2 Extension on encoding of wild cards

2 Fragment Request Unit3 Binary to XML mapping of IPMP-X

E MPEG Multimedia Middleware5678 Reference Software

29116 1 MS MAF ProtocolsJoint management of content description and presentation

X E2E Multimedia PlatformMDS

7 12 Query FormatSchemas

21 4 1 IPMP Components Amendment 18 1 Reference software

IPMP Components DIADIPERFIDDIS

14 ConformanceIPMP Components DIADIPERFID

18 1 Digital Item StreamingSchemas

A 2 Protected Music Player MAF5 Media Streaming MAF6 Professional Archival MAF7 Open Release MAF

Video4 2 4 Simple Profile level 67 3 3 Visual Signature Tools

14

A 3 1 Photo Player Reference Software2 Photo Player Conformance

B 4 Reconfigurable Video CodingC 2 Fixed-point 8x8 IDCT and DCT

4 Reconfigurable Video CodingJVT 4 New AVC Profiles for Professional Applications Conformance

New AVC Profiles for Professional Applications Reference SWScalable Video Coding ConformanceScalable Video Coding Reference SW

10 3 Scalable Video Coding4 Multi-View Video Coding

Audio 2 7 1 Transport of MPEG Surround data in AAC4 3 8 MP4 box for original audio file information

3 9 AAC-ELD3 5 BSAC extensions and transport of MPEG Surround4 14 BSAC conformance

18 MPEG-1 and -2 on MPEG-4 conformance19 ALS conformance20 SLS conformance29 SMR Conformance

5 16 SMR Reference Software15 BSAC and SLS Reference Software

23 SMRA 2 1 Music Player MAF Conformance and reference software

2 Protected Music Player MAF6 Professional Archival MAF

D 1 1 MPEG Surround Reference Software2 MPEG Surround Conformance3 Spatial Audio Object Coding

X Audio and Speech Coding3DG

4 4 16 Conformance MPEG-J GFX21 Conformance of Geometry and shadow

5 11 Reference software MPEG-J GFX13 Reference Software of Geometry and shadow

16 2 Frame-based animated mesh compression3 3D Multiresolution profile

25 3D Graphics Compression modelTest

4 10 3 SVC verification tests ISG

4 9 2 Reference Hardware Description3 Reference Hardware Description

7 6 Reference softwareLiaison

JPEG IPMP-JPSECJPSearch - MP7QF JPSearch – Photo Player MAF

15

10.2 Joint meetings The following joint meetings were held

Groups What Day Where TimeReq. Mds URN, DID, MP7QF Tue Req 09:00-11:00Mds, Sys DI FF issues Tue Mds 11:00-12:00Req, ISG, Vid RVC and AVS Tue Req 12:00-12:30Req, Mds, Sys, Vid, Aud MAFs under cons. Tue Req 14:00-18:00Sys, JPEG JPSEC-IPMP Tue Sys 09:00-10:30?Sys, Aud Mp4 FF Wed Aud 11:30-12:00Req, 3dg 3D compr. Prof. Wed 3dg 12:00-12:30Req, Vid FTV, MVC Wed Jvt 14:00-15:30Vid, Jvt, Req Video metadata carriage Wed Jvt 15:30-16:00Mds, Sys MP21-Laser Wed Sys 16:00-17:00Req, Sys Laser prof. Thu Req 09:00-09:30Vid, JPEG PP MAF, JPSearch Thu Vid 10:00-11:00Mds, JPEG MP7QF Thu Mds 11:00-12:00Mds, Req DID Thu Mds 15:00-16:00

11 Administrative matters

11.1 Schedule of future MPEG meetingsThe following meeting schedule was approved

# City Country yy mm dd-dd80 San José US 07 04 23-2781 Lausanne CH 07 07 02-0682 Shenzhen CN 07 10 22-2683 Antalya TR 08 01 14-1884 Geneva? CH? 08 04-05 28-0285 Hannover DE 08 07 21-2586 Seoul KR 08 10 13-17

11.2 Promotional activities The press release from the 80th meeting was approved

8915 San José press release

12 Planning of future activitiesThe following ad hoc groups were established

9063 Ad Hoc Group on MAF Under Development in Systems

9062 Ad Hoc Group on MPEG File Formats

9061 Ad Hoc Group on Scene Representation

8997 AHG for Video Annotation

9143 AHG on 3DG documents, experiments and software maintenance

16

9097 AHG on Audio Standards Maintenance

8947 AHG on FTV

8990 AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

8996 AHG on MPEG-4 Part 9: Reference Hardware Description Phase 1 and 2.

9128 AHG on MPEG-7 Query Format

8992 AHG on MPEG-7 Visual and Photo Player MAF

8991 AHG on Reconfigurable Video Coding

8946 AHG on Review of MPEG-21 DID

9098 AHG on SAOC CfP, AAC-ELD and Speech and Audio Exploration

8993 AHG on SVC Verification Test

13 Resolutions of this meeting These were approved

14 A.O.B There was no other business

15 ClosingThe meeting closed at 2007/04/27T22:40

17

Annex A – Attendance list

First name Last name Affiliation CountryIan Burnett University of Wollongong AustraliaGerrard Drury University of Wollongong AustraliaMichael Ransburg Klagenfurt University AustriaChristian Timmerer Klagenfurt University AustriaDan Cernea ETRO - VUB BelgiumJan De Cock Ghent University BelgiumSaar De Zutter Ghent University BelgiumRik Van de Walle Ghent University - IBBT BelgiumMichael Gallant LSI Logic CanadaPatrick Rault Quartics CanadaWa James Tam Communications Research Centre Canada

(CRC)Canada

Liang Zhang Communications Research Centre Canada (CRC)

Canada

Weizhong Chen Huawei Technologies Co., Ltd. ChinaQuqing Chen Thomson Broadband R&D (Beijing) Co. Ltd. ChinaLou Dongsheng China Electronics Standardization Institute ChinaYongying Gao Thomson Corporate Research Beijing ChinaWei-Hung Huang MediaTek ChinaYu-Wen Huang MediaTek ChinaJunyan Huo Xidian University ChinaGwo Giun (Chris) Lee National Cheng Kung University ChinaSixin Lin Huawei Tech. Co. Ltd ChinaYang Ping Tsinghua University ChinaHonggang Qi Institute of Computing Technology, Chinese

Academy of SciencesChina

Cliff Reader Self ChinaLianhuan Xiong Huawei Technologies Co., Ltd. ChinaXiaozhong Xu Tsinghua University ChinaHaitao Yang Xidian University ChinaLu Yu Zhejiang University ChinaXiaozhen Zheng Huawei Technologies Co., Ltd. ChinaLihua Zhu Thomson Inc ChinaYing Chen Tampere Univ. Tech. FinlandMiska Hannuksela Nokia FinlandHuopaniemi Jyri Nokia FinlandJani Lainema Nokia FinlandJustin Ridge Nokia FinlandKemal Ugur Nokia FinlandMauri Vaananen Nokia FinlandVincent Bottreau Thomson FranceArnaud Bourge NXP Semiconductors FranceNathalie Cammas Orange-France Telecom R&D FranceSylvain Devillers France Telecom FranceNicolson Didier Thales FranceJulien Dubois University Burgundy FranceJean-Claude Dufourd Streamezzo FrancePatrick Gioia France Telecom FranceMarc Guez Vucher SCPP FranceJoel Jung Orange-France Telecom R&D FranceMohamed-Chaker Larabi SIC, University of Poitiers France

18

Anne LeBris France Telecom FranceKhaled Mammou ARTEMIS Departement Institut National des

TélécommunicationsFrance

Patrice Onno Canon Research Centre France SAS FranceStephane Pateux Orange-France Telecom R&D FrancePierrick Philippe Orange Labs FranceMarius Preda iNT FranceDavid Thevenin Expway FranceJerome Vieron Thomson R&D FranceMatthias Gruhne Fraunhofer IDMT GemanyPeter Amon Siemens AG GermanyGero Bäse Siemens AG GermanyKlaus Diepold Technische Universitaet Muenchen GermanyMario Doeller University of Passau GermanyRalf Geiger Fraunhofer IIS GermanyBernhard Grill Fraunhofer IIS GermanyOliver Hellmuth Fraunhofer IIS GermanyJuergen Herre Fraunhofer IIS GermanyTilman Liebchen LG Electronics GermanyKarsten Müller Fraunhofer HHI GermanyMarkus Multrus Fraunhofer IIS GermanyMatthias Narroschke University of Hannover GermanyJens-Rainer Ohm RWTH Aachen University GermanyJoern Ostermann University of Hannover GermanyThomas Rathgen Ilmenau Technical University GermanyThomas Schierl Fraunhofer IIS GermanyJuergen Schmidt Thomson Inc. GermanyAndreas Schneider Coding Technolegies GmbH GermanyMarkus Schnell Fraunhofer IIS GermanyFlorian Schreiner Technische Universität München GermanyHeiko Schwarz Fraunhofer HHI GermanyAlsosa Smolic Fraunhofer IIS GermanyRalph Sperschneider Fraunhofer IIS GermanyHerbert Thoma Fraunhofer IIS GermanyThomas Wedi Panasonic GermanyMathias Wien RWTH Aachen University GermanySteffen Wittmann Panasonic GermanyPierfrancesco Bellini University of Florence - DISIT-DSI ItalyFilippo Chiariglione CEDEO.net ItalyLeonardo Chiariglione CEDEO.net ItalyGiovanni Cordara Telecom Italia Lab ItalyDavide Rogai University of Florence - DISIT-DSI ItalyKohtaro Asai Mitsubishi Electric Corporation JapanYukihiro Bandoh NTT JapanMark Callow HI Corporation JapanTakeshi Chujoh Toshiba Corporation JapanToshiaki Fujii Nagoya University JapanJunichi Hara Ricoh Company, Ltd. JapanNoboru Harada NTT JapanSatoshi Ito Toshiba Corporation JapanTakashi Itoh Fujitsu Laboratories Ltd. JapanKota Iwamoto NEC Corporation JapanItaru Kaneko Tokyo Polytechnic University JapanHideaki Kimata NTT Corporation JapanTakahiro Kimoto NEC Corporation JapanAbe Kiyofumi Matsushita Electric Industrial Co., Ltd. Japan

19

Takuyo Kogure Matsushita Electric Industrial Co., Ltd. JapanTakehiro Moriya NTT JapanTokumichi Murakami Mitsubishi Electric Corporation JapanJoji Naito JVC JapanSei Naito KDDI Corp. JapanHiroya Nakamura JVC JapanTakashi Nishi Oki Electric Industry Co., Ltd. JapanToshiyuki Nomura NEC Corporation JapanTakeshi Norimatsu Matsushita Electric Industrial Co., Ltd. JapanRyoma Oami NEC Corporation JapanYukiko Ogura IPSJ/ITSCJ JapanSatoru Sakazume Victor Company of Japan, Limited JapanMasanori Sano NHK JapanKazushi Sato Sony Corporation JapanShun-ichi Sekiguchi Mitsubishi Electric Corporation JapanTakanori Senoh National Institute of Info & Comm Tech JapanMasato Shima Texas Instruments Japan JapanOsamu Shimada NEC Corporation JapanShinya Shimizu NTT JapanTaichiro Shiodera Toshiba Corporation JapanKen Sugiyama NEC Corporation JapanTeruhiko Suzuki Sony Corp JapanMasashi Takahashi Hitachi, Ltd JapanSeishi Takamura NTT Cyber Space Laboratories, NTT

CorporationJapan

Masayuki Tanimoto Nagoya University JapanAkiyuki Tanizawa Toshiba Corporation JapanYoichi Yagasaki Sony Corp. JapanAkio Yamada NEC Corporation JapanYoshihisa Yamada Mitsubishi Electric Corporation JapanTomoo Yamakage Toshiba Corporation JapanTomoyuki Yamamoto Sharp Corporation JapanTakahiro Yamasaki Oki Electric Industry Co., Ltd. JapanYoshiyuki Yashima NTT Corporation JapanJeong-Hwan Ahn Samsung Electronics KoreaSunguk Baik Oniontech co.,ltd KoreaHyouk Jean Cha LG Electronics KoreaJihun Cha ETRI KoreaSeo Chanwon Sejong University KoreaAyoung Cho Inha University KoreaByeongho Choi KETI KoreaHae Chul Choi ETRI KoreaMiran Choi ETRI KoreaWoong Il Choi Samsung KoreaYungho Choi SK Telecom KoreaJong Bum Choi Samsung Electronics KoreaHyon-Gon Choo ETRI KoreaHyon-Gon Choo ETRI KoreaSung-Moon Chun ECT Inc. KoreaWoo-Jin Han Samsung Electronics KoreaKi Hun Han Sejong University KoreaMin Cheol Hong Soongsil University KoreaChi Jung Hwang ChungNam Univ KoreaLee James KETI KoreaEuee Seon Jang Hanyang University KoreaByeong Moon Jeon LG Electronics Korea

20

Byeungwoo Jeon SKKU KoreaYongjoon Jeon LG Electronics KoreaDongseak Jeong Inha University KoreaJie Jia Sejong University KoreaSung Ho Jin Information and Communications University KoreaSanhhyun Joo ETRI KoreaYe Sun Joung ETRI KoreaJaebum Jun Hanyang University KoreaYang-Won Jung LG Electronics KoreaBongsoo Jung SKKU KoreaJung Won Kang ETRI KoreaChang Ick Kim Information and Communications University KoreaDo-Hyung Kim Samsung Advanced Institute of Technology KoreaDong Soo Kim LG Electronics KoreaHae Kwang Kim Sejong University KoreaHui Yong Kim ETRI KoreaHyun Mun Kim Samsung AIT KoreaHyungyu Kim Hanyang University KoreaJae-Gon Kim Hanbat National University KoreaJingwoong Kim ETRI KoreaJong Lak Kim DSP Group KoreaMunchurl Kim Information and Communications University KoreaSo Young Kim Samsung Electronics KoreaTae Hyeon Kim LG Electronics KoreaTaehyun Kim DRM inside KoreaYong Goo Kim SK Telecom KoreaYong Han Kim University of Seoul KoreaYong-Hwan Kim KETI KoreaDongkyun Kim Sejong University KoreaJae-Il Kim Information and Communications University KoreaHan-Suh Koo LG Electronics KoreaSang Hoon Lee DSP Group KoreaSangHeon Lee Seoul Nat'l Univ KoreaSun Young Lee Hanyang University KoreaYung Lyul Lee Sejong University KoreaChungku Lee HUMAX Co.,Ltd. KoreaSangyoun Lee Yonsei University KoreaYungKi Lee Sejong University KoreaSungChang Lim Sejong University KoreaYoung-Kwon Lim net&tv Inc. KoreaTaebeom Lim Korea Electronics Technology Institute KoreaMoon Nam Mee Seoul University of Venture & Information KoreaHack Youp Noh Korea KoreaWeongeun Oh ETRI KoreaHenney Oh LG Electronics KoreaKwan-Jung Oh GIST (Gwangju Institue of Science and

Technology)Korea

Hee-Suk Pang LG Electronics KoreaJi Ho Park KETI KoreaMin Woo Park Kyung Hee University KoreaSeanae Park Kwangwoon Univ. KoreaSeung-Wook Park LG Electronics KoreaDongHwan Park LG Electronics KoreaMuhammad Syah Houari

Sabirin Information and Communications University Korea

Jeongil Seo ETRI Korea

21

Jungdong Seo Yonsei University KoreaHee-Cheol Seo ETRI KoreaJuheon Seo Sejong University KoreaWoo Sung Shim Samsung Electronics KoreaDonggyu Sim Kwangwoon Univ. KoreaJaeyeon Song Samsung Elecronics KoreaDoug Young Suh KHU KoreaJong-Yeul Suh LG Electronics KoreaHendry Tan Information and Communications University KoreaJungyoup Yang Sungkyunkwan University KoreaJeong-Hyu Yang LG Electronics KoreaWon Keun Yang ETRI KoreaAlex Chungku Yie HUMAX Co.,Ltd. KoreaJeong-ju Yoo ETRI KoreaJisang Yoo Kwangwoon University KoreaKyoungro Yoon Konkuk University KoreaSungyong Yoon LG Electronics KoreaJianhua Zheng Huawei Technologies Co., Ltd. KoreaSebastien Brangoulo Joost Technologies NetherlandsJeroen Breebaart Philips Research NetherlandsFons Bruls Philips NetherlandsJean H.A. Gelissen Philips Research NetherlandsJohan Muskens Philips Research NetherlandsWerner Oomen Philips Applied Technologies NetherlandsGisle Bjøntegaard Tandberg NorwayMarian Muczko Telekomunikacja Polska PolandLukasz Pikula Telekomunikacja Polska PolandFernando Pereira IST-IT PortugalKok Seng Chong Panasonic Singapore Laboratories SingaporeKwong Huang Goh Institute for Infocomm Research SingaporeHaibin Huang Institue for Infocomm Research SingaporeKelvin Lee Institute for Infocomm Research SingaporeTe Li Institute for Infocomm Research SingaporeZhengguo Li Institute for Infocomm Research SingaporeChong Soon Lim Panasonic Singapore Laboratories SingaporeLeong Mun Kew Institute For Infocomm Research SingaporeSua Hong Neo Panasonic Singapore Laboratories SingaporeSusanto Rahardja Institute for Infocomm Research SingaporeShengmei Sheng Panasonic Singapore Labs SingaporeWei Yao Institute for Infocomm Research SingaporeThiow Keng Tan NTT DoCoMo, Inc. SingaporeJaime Delgado Universitat Politècnica de Catalunya SpainRuben Tous Universitat Politecnica de Catalunya SpainPer Fröjdh Ericsson SwedenKristofer Kjörling Coding Technologies AB SwedenHeiko Purnhagen Coding Technologies AB SwedenJonas Rödén Coding Technologies AB SwedenRickard Sjöberg Ericsson SwedenPeirre Davy University of Geneva SwitzerlandTouradj Ebrahimi EPFL SwitzerlandChristophe Lucarz EPFL SwitzerlandMarco Mattavelli EPFL SwitzerlandTanya Beech QinetiQ UKMiroslaw Bober Mitsubishi Electric Corporation UKLeszek Cieplinski Mitsubishi Electric ITE-VIL UKKate Grant Nine Tiles UK

22

Mike Nilsson BT UKRobert O'Callaghan Mitsubishi Electric ITE-VIL UKPing Wu Tandberg Television UKJose Roberto Alvarez Mobilygen Corporation USAYiliang Bao Qualcomm USALazar Bivolarski BrightScale, Inc USAPeter Borgwardt Motorola USAMadhukar Budagavi Texas Instruments Inc. USAWo Chang NIST USALulin Chen Omneon Video Networks USAYi-Jen Chiu Intel Corp. USAHyukjune Chung Qualcomm Inc. USAReha Civanlar DoCoMo USA Labs USAKatie Cornog Avid Technology USAGuy Cote Mobilygen Corporation USAOscar Divorra

EscodaThomson Inc. USA

James Durham JITC / DISA USAAlex Eleftheriadis Layered Media, Inc. USAMatt Fellers Dolby Laboratories USAOnur Guleryuz DoCoMo USA Labs USAOztan Harmanci DoCoMo USA Labs USABarry Haskell Apple Inc. USAPaul Haskell Harmonic, Inc. USAJones He Freescale Semiconductor USAArianne Hinds IBM USADanny Hong Layered Media, Inc. USAShih-Ta Hsiang Motorola USAWalt Husak Dolby Laboratories USAFaisal Ishtiaq Motorola USAMichael Isnardi Sarnoff Corporation USAJorn Janneck Xilinx USASandeep Kanumuri DoCoMo USA Labs USAMukta Kar CableLabs USAMarta Karczewicz Qualcomm USAJae Hoon Kim University of Southern California USAArkady Kopansky Sarnoff Corporation USAShawmin Lei MediaTek USAAthanasios Leontaris Dolby Laboratories USAVladimir Levantovsky Monotype Imaging Inc. USAHe-Yuan Lin NCKU USAYuxin Liu Hewlett Packard Company USANing Lu Intel USAJiancong Luo Thomson Inc. USAAjay Luthra Motorola USASean McCarthy Modulus Video USAJim Meany Boeing USADebargha Mukherjee Hewlett Packard Company USASam Narasimhan Motorola USAObianuju Ndili Santa Clara University USATokunbo Ogunfunmi Santa Clara University USAPurvin Pandit Thomson Inc. USAWen-Hsiao Peng National Chiao-Tung University/ITRI USAYolanda Prieto Freescale Semiconductor USASchuyler Quackenbush Audio Research Labs USAShankar Regunathan Microsoft Corporation USA

23

Yuriy Reznik Qualcomm Inc. USACharles Robinson Dolby Laboratories USAArturo Rodriguez Cisco USAMike Rubinfeld NIST USAJesus Sampedro Polycom, Inc. USAAndrew Segall Sharp USAXiaojin Shi Apple Inc. USAPrasanna Singamsetty Intel Corporation USADavid Singer Apple USARamin Soheili Seda Solutions Corporation USAYeping Su Sharp Labs of America USAGary Sullivan Microsoft Corporation USAHuifang Sun Mitsubishi Electric Research Labs USAYasser Syed Hewlett Packard Company USAAli Tabatabai Sony USAAndrew Tescher Microsoft Corporation USADong Tian Thomson Inc. USAPankaj Topiwala FastVDO USAAlexandros Tourapis Dolby Laboratories USAChun-Jen Tsai NCTU/ITRI USAYi-Shin Tung Setabox Technology Corporation USAJames Van Loo Microsoft Corporation USAVictor Vedovato Microsoft Corporation USAAnthony Vetro Mitsubishi Electric Corporation USAEric Viscito eV Consulting USAMohammed Visharam Sony USAWade Wan Broadcom Corporation USAHaohong Wang Marvell Semiconductors USAXianglin Wang Nokia USAYong Wang Motorola USAXin Wang ContentGuard, Inc. USAHitoshi Watanabe Qpixel Technology, Inc. USASamuel Wong Intel USAHsi-Jung Wu Apple Inc USAJohn Wus Panasonic USAYan Ye Qualcomm Inc USAPeng Yin Thomson USAHaoping Yu Thomson Inc. USASheng Zhong Broadcom Corporation USA

24

Annex B – Agenda

      Item1    Opening 2    Roll call of participants 3    Approval of agenda4    Allocation of contributions 5    Communications from Convenor6    Report of previous meeting 7    Processing of NB Position Papers8 1  Media coding    1Fixed point implementation of DCT/IDCT    2Advanced 4:4:4 Profile    3Scalable Video Coding    4Multiview Video Coding    5BSAC Extensions    6MPEG Surround    7Geometry and Shadow     8Reconfigurable Video Coding    9Video Tool Library    10Scalable audio and speech coding  2  Composition coding    1Lightweight Scene Representation    2Symbolic Music Representation  3  Description Coding    1Schema definition    2Visual Descriptor Extensions    3MPEG-7 Query Format  4  Systems support    1Fragments Request Unit    2 JPEG2000 support in MPEG-4 Systems  5   IPMP    1MPEG-21 IPMP Component Base Profile    2REL Profiles  6  Digital Item    1Schema files for MPEG-21 standards  7 1Transport and File Format    2Transport of Auxiliary Video Data    3Transport of MPEG Surround data in AAC    4File Format extensions for Description of Timed Metadata    5Flute Hint Track    6AVC File Format extensions for FRExt    7AVC File Format extensions for SVC    8File Format Issues for Support of Audio Media    9Digital Item Streaming

25

  8  Multimedia architecture    1M3W Component Download    2M3W Fault Management    3M3W System Integrity Management    4M3W Reference Software  9  Application formats    1Protected Music Player MAF    2Photo Player MAF    3Musical Slide Show MAF    4Media Streaming MAF           5Professional Archival MAF    6Open Release Application Format    7Portable Video Player    8Digital Multimedia Broadcasting Application Format    9Exploration  10  Reference implementation    1File Format Reference Software    2Reference Hardware Description    3MPEG Surround Reference Software    4Symbolic Music Representation    5Morphing & Textures Reference Software    6MPEG-J GFX Reference Software    7MPEG-7 Systems Reference Software    8Perceptual 3D Shape Reference Software    9MPEG-21 REL Reference Software    10MPEG-21 DIA Reference Software    11Binary MPEG format for XML Reference Software    12Prefixes and wild card extensions reference software     13M3W Reference Software  11  Conformance    1Audio BIFS v3 Conformance    2MPEG-1 and -2 Audio in MPEG-4 Conformance    3BSAC conformance    41-bit Oversampled Audio Conformance    5Audio Lossless Conformance    6Audio Scalable to Lossless conformance    7MPEG Surround conformance    8Symbolic Music Representation    9Morphing & Textures Conformance    10File Format conformance    11Advanced Text and Graphics Conformance    12MPEG-J GFX Conformance    13Open Font Format conformance    14Perceptual 3D Shape Conformance    15 IPMP Components Conformance

26

    16Event Reporting Conformance    17Fragment Identification of MPEG Resources Conformance    18Music Player Application Format Conformance    19Binary MPEG format for XML Conformance    20Prefixes and wild card extensions conformance     21M3W Conformance  12  Maintenance    1Systems coding standards    2Video coding standards    3Audio coding standards    4Visual description coding standards    5Audio description coding standards    6MDS standards

9     Liaison matters 10     Organisation of this meeting       Tasks for subgroups      Joint meetings 11     Administrative matters      Schedule of future MPEG meetings       Promotional activities 12     Planning of future activities13     Resolutions of this meeting 14     A.O.B 15     Closing

27

Annex C – Input contributions

No. Authors Title

14268 Wo Chang Document Register for SC29/WG11 Meeting San Jose, USA

14269Francisco Mor. Burgos (UPM)Jeong-Hwan AhnMark Callow

AHG on 3DG documents, experiments and software maintenance

14270

Marco MattavelliG. SullivanA. HindsY. ReznikP. Topiwala

AHG on Video IDCT Specification

14271 Yi-Shin TungChung-Neng Wang

AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

14272 Euee S. JangYoshihisa Yamada AHG on Reconfigurable Video Coding

14273Sang-Kyun KimRobert O'CallaghanAkio Yamada

AHG on Maintenance of MPEG-7 Visual related Documents, Reference Software and Conformance

14274

Miroslaw BoberSang-Kyun KimAkio YamadaWo Chang

AHG on MPEG-7 Visual and Photo Player MAF

14275 Wo Chang AHG on MAFs Awareness Event

14276 Robert TurneyMarco Mattavelli

AHG on MPEG-4 Part 9 Reference Hardware Description Phase 2 and 3

14277 Gerrard DruryPeder Drege AHG on MPEG-21 DIS

14278Filippo ChiariglioneChristian TimmererThomas Skjolberg

AHG on the Media Streaming MAF demo for the MAF-AE

14279Stefan KraegelohFilippo Chiariglione Noboru Harada

AHG on MDS MAFs Under Development

14280Wo ChangKyoungro YoonMario Doeller

AHG on MPEG-7 Query Format

14281 R. Sperschneider AHG on Audio Standards Maintenance

14282 S. Quackenbush AHG on SAOC CfP and AAC-ELD

14283 Tobias OelbaumMathias Wien

AHG on SVC Verification Test

28

Justin RidgeVincent BottreauNathalie CammasAlexandros Eleftheraidis

14284 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-5:2001/FDAM 10 [SC 29 N 8174]

14285 W3C via SC 29 Secretariat Liaison Statement from W3C [SC 29 N 8177]

14286 SC 29 Secretariat Summary of Response to Proposal of Minor Enhancement: 14496-3/Amd.9 [SC 29 N 8179]

14287 SC 29 Secretariat Summary of Voting on ISO/IEC TR 11172-5:1998/DCOR 1 [SC 29 N 8178]

14288 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2005/PDAM 9 [SC 29 N 8180]

14289 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 24 [SC 29 N 8182]

14290 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 25 [SC 29 N 8184]

14291 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-5:2004/PDAM 3 [SC 29 N 8190]

14292 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-3:2005/FDAM 1 [SC 29 N 8207]

14293 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 13818-1:200X/FDAM 1 [SC 29 N 8211]

14294 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-5:2003/PDAM 3 [SC 29 N 8212]

14295 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-7:2003/PDAM 4 [SC 29 N 8213]

14296 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23002-2 [SC 29 N 8222]

14297 3GPP via SC 29 Secretariat Liaison Statement from 3GPP [SC 29 N 8225]

14298 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23000-5 [SC 29 N 8226]

14299 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23001-3 [SC 29 N 8227]

14300 FG IPTV via SC 29 Secretariat Liaison Statement from ITU-T IPTV Focus Group (FG IPTV) [SC 29 N 8228]

14301

Christophe Lucarz Marco Mattavelli Andrew KinaneSunyoung Lee Sinwook Lee

RVC Functional Units naming process proposal

14302 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC FDIS 14496-22 [SC 29 N

29

8234]

14303 SC 29 Secretariat Summary of Voting on NWIP, Information technology -- Supplemental media technologies [SC 29 N 8235]

14304 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23005-1 [SC 29 N 8236]

14305 the DVD Forum WG-1 via SC 29 Secretariat

Liaison Statement from the DVD Forum WG-1 [SC 29 N 8254]

14306 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-4:2004/FDAM 12 [SC 29 N 8249]

14307 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-5:2001/FDAM 9 [SC 29 N 8251]

14308 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-11:2005/FDAM 5 [SC 29 N 8252]

14309 SC 29 Secretariat Summary of Voting on ISO/IEC 23002-1/PDAM 1 [SC 29 N 8259]

14310Yuriy A. ReznikGary SullivanArianne T. Hinds

Study Text of ISO/IEC 23002 CD (editors input)

14311 Yuriy Reznik Study Text of ISO/IEC 23002-1/PDAM1 (editors input)

14312 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 13818-2:2000/FDAM 2 [SC 29 N 8260]

14313 IEC TC 100 via SC 29 Secretariat IEC CDV 61937-3 [SC 29 N 8263]

14314 IEC TC 100 via SC 29 Secretariat IEC CDV 61966-2-5 [SC 29 N 8264]

14315 Schuyler Quackenbush Spatial Audio Object Coding Evaluation Procedures and Criterion

14316 Schuyler Quackenbush 79th MPEG Audio Report

14317 Schuyler Quackenbush Proposed Workplan for Speech and Audio Exploration

14318 Sylvain Devillers Editors' input to draft text of 23001-5 (MPEG-B BSDL)

14319 SC 29 Secretariat Summary of Voting on ISO/IEC 13818-7:2006/FPDAM 1 [SC 29 N 8268]

14320 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 18 [SC 29 N 8269]

14321 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 19 [SC 29 N 8270]

14322 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 20 [SC 29 N 8271]

14323 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 21 [SC 29 N 8272]

14324 SC 29 Secretariat Summary of Voting on ISO/IEC

30

14496-5:2001/FPDAM 12 [SC 29 N 8273]

14325 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-6:2003/FPDAM 2 [SC 29 N 8274]

14326 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-7:2003/FPDAM 3 [SC 29 N 8275]

14327 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 14 [SC 29 N 8276]

14328 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 14496-23 [SC 29 N 8277]

14329 A. G. Tescher for USNB USNB Contribution: Response to resolution 3.1.2 of 79-th WG 11 meeting

14330

Thomas SkjølbergPeder DregeJoseph Thomas-KerrGerrard Drury

Report of CE on DIS TuC

14331 ETSI TC DECT via SC 29 Secretariat

Liaison Statement from ETSI TC DECT to ITU-T SG 12 and ETSI TC STQ

14332 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-5:2001/PDAM 13 [SC 29 N 8280]

14333 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-12:2005/FDAM 1 [SC 29 N 8281]

14334 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-4:2006/FPDAM 1 [SC 29 N 8282]

14335 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-18/PDAM 1 [SC 29 N 8294]

14336 SC 29 SecretariatSummary of Voting on ISO/IEC 14496-12:2005/FPDAM 2 and ISO/IEC 15444-12:2005/FPDAM 2 [SC 29 N 8297]

14337 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23004-5 [SC 29 N 8298]

14338 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23004-6 [SC 29 N 8299]

14339 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23004-7 [SC 29 N 8301]

14340 Christophe LucarzMarco Mattavelli

Compression of the RVC DDL Decoder Description with BiM (results of Core Experiment 1.3 in RVC)

14341Christian TimmererSylvain DevillersMichael Ransburg

Editor's input on Draft MPEG-21 DIA 2nd edition

14342 CEA via SC 29 Secretariat Liaison Statement from CEA [SC 29 N 8310]

14343 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23000-4 [SC 29 N 8306]

31

14344 SC 29 Secretariat Summary of Voting on ISO/IEC 23003-1/PDAM 1 [SC 29 N 8307]

14345 SC 29 Secretariat Summary of Voting on ISO/IEC 23003-1/PDAM 2 [SC 29 N 8308]

14346 Yuriy Reznik Updated 23002-1 IDCT precision testbed

14347 Yuriy ReznikArianne Hinds Updated H.263-based IDCT testbed

14348 Arianne T. Hinds Updated MPEG-4 IDCT Testbed

14349Gavin SchutzTeruhiko SuzukiMichael Dolan

Liaison re w8559 Text of ISO/IEC 13818-1:200x/DCOR.1

14350

Weon-Geun OhDong-Seok JeongJu-Kyoung JinA-Young ChoJun-Woo LeeIk-Hwan ChoWon-Keun Yang

Mathematical consideration on the degree of geometrical modification

14351Saar De ZutterJan De CockRik Van de Walle

Conformance tests for DIDL documents - files

14352 James Orwell Contribution to the Basic Video Surveillance MAF

14353 ATIS IIF via SC 29 Secretariat Liaison Statement from ATIS IIF [SC 29 N 8317]

14354 ITU-T SG 16 via SC 29 Secretariat Liaison Statement from ITU-T SG 16 [SC 29 N 8324]

14355 Ralph Sperschneider WD on MPEG-4 Audio Fourth Edition

14356

Saar De ZutterJan De CockRik Van de Walleon behalf of the Belgian National Body

BNB comments on ISO/IEC FCD 21000-14: Conformance Testing

14357 jungwonLee ISO/IEC JTC 1/SC 29/WG 11 N6702

14358 Yi=Shin TungJa-Ling Wu

Additional fixes on MPEG-4 video conformance bitstreams

14359 Yi-Shin TungJa-Ling Wu Consider row-transform-first IDCT in 23002-2

14360 A. G. Tescher for USNB USNB Contribution: Issues relating to expiring patents

14361 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 21000-14 [SC 29 N 8332]

14362 DVB via SC 29 Secretariat Liaison Statement from the DVB [SC 29 N 8326]

14363 Nikolce StefanoskiJörn Ostermann Scalable Compression of Dynamic 3D Meshes

14364 Pierfrancesco Bellini Editors study on ISO/IEC 14496-23/FCD

32

Paolo NesiMaurizio CampanaiGiorgio Zoia

14365Davide RogaiPaolo NesiPierfrancesco Bellini

Experience on using MPEG-21 File Format for nested and/or protected DIs

14366Paolo NesiPierfrancesco Bellini Davide Rogai

Additional examples on Cross-Media Interactive Presentation MAF

14367

Paolo NesiPierfrancesco Bellini Davide RogaiKia Ng (University of Leeds)

Proposal for a MAF on Cross-Media Interactive Presentation: Application Scenarios

14368Paolo NesiPierfrancesco Bellini Davide Rogai

Proposal for a MAF on Cross-Media Interactive Presentation: Requirements

14369Davide RogaiPierfrancesco BelliniPaolo Nesi

Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFs

14370 Jean-Claude Dufourd LASeR fixes requested by 3GPP DIMS

14371 Jean H.A. Gelissen (editor)Johan Muskens

Contribution to M3W Reference Software for M3W Parts 2, 3, 5, 6 & 7

14372 Jean-Claude Dufourd Splitting LASeR AMD1

14373 Jean-Claude Dufourd LASeR profiles adjustments

14374Gwo Giun LeeHe-Yuan LinMing-Jiun Wang

Functional units of inter-prediction under reasonable system partition for RVC framework

14375Gwo Giun LeeHe-Yuan LinMing-Jiun Wang

Conformance test tools of RVC functional units

14376 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23000-2

14377 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 16

14378 Jean-Claude Dufourd Additions to LASeR AMD2 from 3GPP

14379 Arianne T. Hinds Updated T.83 testbed for IDCTs

14380 Zhibo Ni Updated MPEG-2 IDCT Testbed

14381 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 23001-2

14382 SC 29 Secretariat Summary of Voting on ISO/IEC 13818-1:200X/DCOR 1

14383 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-2:2004/PDAM 4

14384 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2005/PDAM

33

8

14385 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 23

14386 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 28

14387 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-11:2005/DCOR 6

14388 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-12:2005/DCOR 3 & ISO/IEC 15444-12:2005/DCOR 3

14389 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-21:2006/DCOR 1

14390 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-3:2002/Amd.2:2006/DCOR 1

14391 SC 29 Secretariat Summary of Voting on ISO/IEC 15938-6:2003/Amd.1:2006/DCOR 1

14392 SC 29 Secretariat Summary of Voting on ISO/IEC 21000-7:2004/DCOR 1

14393 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23000-7

14394 SC 29 Secretariat Summary of Voting on ISO/IEC CD 23000-9

14395 SC 29 Secretariat Summary of Voting on ISO/IEC 23001-1:2006/DCOR 2

14396 Jeong-Hwan Ahn Conformance bitstream for Geometry & Shadow

14397 SC 29 Secretariat Late Vote on ISO/IEC 21000-4:2006/FPDAM 1 [SC 29 N 8331]

14398 SC 29 SecretariatCommon Patent Policy for ITU-T/ITU-R/ISO/IEC, and Guidelines for Implementation of the Common Patent Policy [SC 29 N 8314]

14399 Eva RodríguezJaime Delgado

Adding Integrity and authenticity to Event Reporting information

14400 Jaime DelgadoEva Rodríguez Defect Report Proposal of ISO/IEC 21000-15

14401 Eva RodríguezJaime Delgado Contribution to REL MAM Profile Conformance

14402 Simon DanielsVladimir Levantovsky

Proposed conformance test methodology and bitstreams for ISO/IEC 14496-22

14403 Arianne T. Hinds Updated TM5 MPEG-2 Testbed

14404 Jani PeltotaloMiska M. Hannuksela

Comments and suggestions regarding ISO/IEC 14496-12 Amd.2

14405 David Singer Comments on the SVC File Format

14406 Sangki KimHyobin Lee

CE Report for VCE-5

34

Sangyoun Lee

14407 Kelvin Lee Status of SLS reference software update

14408

Marius Preda Benoit Le Bonhomme Son Tran Françoise Preteux

3dod.org goes multimedia: MyMultimediaWorld.com

14409

Saar De ZutterJan De CockRik Van de Walleon behalf of the Belgian National Body

Preliminary BNB comments on ISO/IEC FCD 21000-8: Reference Software (2nd edition)

14410Noboru HaradaTakehiroMoriyaYutaka Kamamoto

Proposed revision for ISO/IEC14496-3, AMD8: MP4FF box for original audio file information

14411Noboru Harada Takehiro Moriya Yutaka Kamamoto

Proposed text to WD of Professional Archical MAF

14412Weon-Geun OhWon-Keun YangDong-Seok Jeong

Modified GST Based Descriptor for MPEG-7 VCE-6 Complex Condition

14413 TTA via SC 29 Secretariat Liaison Statement from TTA [SC 29 N 8333]

14414Kelvin LeeTe LiHaibin Huang

Proposed Corrigenda to 14496-3:2005/AMD 3 (SLS)

14415Kisong YoonTaehyun KimHogab Kang

Interoperability between MPEG-21 REL DAC Profile and Other Standards

14416 Jar-Sheng ChenChun-Jen Tsai

Implementation of B frame support in RVC CAL Model

14417

Masayuki TanimotoToshiaki FujiiHideaki KimataShigeyuki Sakazawa

Proposal on Requirements for FTV

14418

Jihun ChaYeSun JoungYoung-Kwon LimKyungAe Moon

Ideas on MPEG-21 and LASeR

14419

Jihun ChaYoungkwon LimYeSun JoungKyungAe Moon

Issues on the carriage of ISO/IEC 14496-20 contents over MPEG-2

14420 Hee-Cheol SeoMiran ChoiHyunki KimMyung-Gil Jang

CE Report for Query Expression of MPEG-7 Query Format

35

Soojong LimJeong HeoKyoungro Yoon

14421

Hee-Cheol SeoMiran ChoiHyunki KimMyung-Gil JangSoojong LimJeong HeoKyoungro Yoon

Revision of Proposed Input Query Format for MPEG-7 Query Format

14422

Hyun-Kook LeeHee-Suk PangDong Soo KimSung-Yong YoonHenney OhYang-Won Jung

Report on the SAOC test material provided by LGE

14423

Kwangcheol ChoiSung-Moon ChunJaedo KwakSeungheon YangJi-Sang YooSi-Hun SungSeong-Cheol Han

Requirements for Stereoscopic MAF

14424

Jaedo KwakSi-Hun SungSung-Moon ChunJinWoong KimNamho Hur

Whitepaper of Stereoscopic Project

14425Hui Yong KimHyon-Gon Choo Munchurl Kim

(Editors Input) Updated Text of ISO/IEC 23000-9 MAF for DMB

14426

Hui Yong KimGun BangMyungSeok KiHyun Cheol KimHan-Kyu LeeJin Woo HongYoung-Kwon Lim

Proposal for MPEG-2 TS Encapsulation with ISO/IEC 23000-9 MAF for DMB

14427

Hui Yong KimSeung Jun YangHeekyung LeeHan-Kyu LeeJin Woo HongMunchurl KimJinhan Kim

Proposal for Restrictions on TV-Anytime Metadata in ISO/IEC 23000-9 MAF for DMB

14428 Tilman Liebchen Proposed Text of ISO/IEC 14496-4:2004/FDAM 19, Audio Lossless Coding (ALS) Conformance

36

14429 Tilman Liebchen Updated Status of ALS Conformance

14430 Tilman Liebchen Comments on Professional Archival MAF Requirements

14431Yo-Sung HoCheon LeeKwan-Jung Oh

CE6: View Interpolation Prediction for Multi-view Video Coding

14432Yo-Sung HoKwan-Jung OhCheon Lee

Observations of Multi-view Test Sequences

14433Yo-Sung HoKwan-Jung OhCheon Lee

CE5: Verification of JVT-W031

14434

Julien DuboisBarthelemy HeyrmanMarco MattavelliJohel Miteran

Wildcard Platform Vs ML310

14435Hyouk Jean ChaTae Hyeon KimHerbert Thoma

Proposed text of ISO/IEC 23000-8 CD Portable video player MAF

14436 Ryoma Oami CE report for VCE-3 on person identity-based photo indexing

14437 Ryoma Oami A proposal on metadata modification for Musical Slide Show MAF

14438 Ryoma Oami A proposal of an additional functionality to be supported in Portable Video Player MAF

14439 Kota IwamotoRyoma Oami CE report for VCE-7 on video signature

14440 Kota IwamotoRyoma Oami Proposal of CE procedure for VCE-7

14441Oliver HellmuthJuergen HerreThorsten Kastner

Proposed SAOC test items provided by Fraunhofer IIS

14442 Hyon-Gon ChooFilippo Chiariglione

Proposed text of ISO/IEC 23000-5 FCD Media Streaming MAF

14443Filippo Chiariglione(Editor)Hyon-Gon Choo(Editor)Jooyoung Lee

Proposed text of ISO/IEC 23001-3 FCD Binary XML to IPMP-X

14444Hyon-Gon ChooFilippo ChiariglioneNaito Joji

Proposed text of ISO/IEC 23005-1 FCD Media Streaming MAF Protocol (Editor's Input)

14445Giseok SonSinwook LeeEuee S. Jang

Core Experiment Result on CDDL

37

14446 Hyungyu KimEuee S. Jang Proposed Text of RVC CE

14447Jaebum JunSunyoung LeeEuee S. Jang

Study on RVC Framework and Its Requirements

14448Yoshihisa YamadaKenji OtoiKohtaro Asai

Proposed text of the RVC FUs for MPEG-4 AVC (Results of CE 2.2)

14449DoellerGruhne Wolf

MP7QF CE Test Report

14450 David Thevenin Philippe de Cuetos Editor's study of 23001-1 FPDAM2

14451 David Thevenin Philippe de Cuetos Binary Conformance streams for MPEG-21

14452Tokumichi MurakamiKohtaro AsaiYoshihisa Yamada

Requirement of Full-Color Video Coding for Consumer Applications

14453

Kristofer KjörlingJonas RödénJeroen KoppensErik SchuijersJeroen Breebaart

Proposed draft corrigendum for MPEG Surround

14454 Christophe LucarzMarco Mattavelli

Implementation of multiple reference frame support in RVC CAL model

14455 Eunmi Oh Evaluation of speech and audio coding scheme

14456Christian TimmererHermann Hellwagneron behalf of Austrian NB

Austrian NB comments on ISO/IEC 21000-14 FCD

14457

Ghislain RoquierMaxime PelcatMickaël RauletMatthieu WipliezJean-François NezanOlivier Déforges

A scheme for implementing MPEG-4 SP codec in the RVC framework

14458

Ingo KoflerChristian TimmererHermann Hellwagneron behalf of Austrian NB

Austrian NB comments on ISO/IEC 21000-7 Cor.1

14459

Michael EberhardChristian TimmererHermann Hellwagneron behalf of Austrian NB

Austrian NB comments on ISO/IEC 21000-8 FCD

14460 Christian TimmererHermann Hellwagner

Austrian NB comments on ISO/IEC CD XXXXX Media Streaming MAF Protocols

38

14461Christian TimmererMichael RansburgHermann Hellwagner

Austrian NB comments on ISO/IEC 23000-5 CD

14462Michael EberhardChristian TimmererHermann Hellwagner

Update of gBSDtoBin and DIA Reference and Utility Software Modules

14463

Maxime PelcatMédéric BlestelMickaël RauletJean-François NezanOlivier Déforges

Evolutions of RVC so as to handle SVC decoding

14464 Jeroen BreebaartWerner Oomen Proposed SAOC test items provided by Philips

14465 Erik SchuijersWerner Oomen Crosscheck FT enhanced LD AAC core experiment

14466Patrick GioiaAnne Le BrisRomain Cavagna

Report of CE2: Space Partitioning

14467 Patrick GioiaOlivier Aubault Proposal for 3D Compression Profile

14468

Nicola AdamiRiccardo LeonardiPierangelo MiglioratiClaudia Tonoli

Performance of a Distributed Video Codec in Presence of Transmission Errors

14469

Honggang QiWen GaoDebin ZhaoSiwei Ma

Crosscheck for IDCT CD

14470 Paul BrasnettMiroslaw Bober Improved Image Identifier (VCE6)

14471 Paul BrasnettMiroslaw Bober Modification of VCE6 Experimental Conditions

14472 Paul BrasnettMiroslaw Bober VCE7 Experimental Conditions

14473

Honggang QiWen GaoTiejun HuangLu Yu

Extension to support non-MPEG standards (ICT/ZJU) (Results of CE 1.6)

14474

Honggang QiWen GaoLu YuEuee S. JangMarco MattavelliAndrew Kinane

Exploration experiments of AVS decoder description in RVC framework

14475 Giovanni Cordara (on behalf of the Italian NB proposal to revisit MPEG-21 DID

39

ITNB)

14476 AVS Workgroup Liaison Statement to MPEG on RVC

14477HendryHouari SabirinMunchurl Kim

Updated Proposal for Protected Musical Slide Show MAF with IPMP

14478HendryHouari SabirinMunchurl Kim

Updated Proposal for Protected Photo Player MAF with IPMP

14479

Taehyun KimJaime DelgadoFlorian SchreinerChris Barlas

Editor's study of ISO/IEC 21000-5/PDAM3

14480 Paul Schumacher Implementation of MPEG-4 AVC Deblocking Filter in RVC CAL model

14481 HendryTakafumi Ueno

Some Editorial Update for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14482 Hendry Late comment for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14483 HendryMunchurl Kim

Contribution for MPEG-21 IPMP Components Base Profile Conformance

14484Kisong YoonTaehyun KimHogab Kang

A Study on Use Cases of Derivative Works with MPEG-21 REL ORC Profile License

14485 Zhibo NiLu Yu IDCT Core Experiment Results

14486Houari SabirrinJeongyeon LimMunchurl Kim

A Proposal for Basic Video Surveillance Application Format

14487 Michael RansburgHermann Hellwagner

Contribution to Conformance for ISO/IEC 14496-12 AMD/1

14488 Jonas EngdegårdBarbara Resch

Description of SAOC test items provided by Coding Technologies

14489Filippo ChiariglioneJooyoung LeeHyon-Gon Choo

Proposal of Modified IPMP XML messages for ISO/IEC 23001-3 Binary XML to IPMP-X

14490

Christophe Lucarz Marco Mattavelli Joseph Thomas-Kerr Jörn Janneck

Reconfigurability potential of the MPEG-4 SP decoder (results of CE 1.1)

14491

Khaled MamouMarius PredaTitus ZahariaFrancoise Prêteux

FAMC bitstream description

40

14492 Fredrik Henn Cross check of FhG Core Experiment on LD-SBR filterbank for AAC-ELD

14493

Khaled MamouKarsten MüllerDetlev MarpeTitus ZahariaMarius PredaFrancoise Prêteux

Frame-based Animated Mesh Compression : integration of the CABAC arithmetic encoder

14494

Thomas RathgenMichael RansburgPeter AmonAndreas HutterHermann Hellwagner

Extraction path description

14495

Michael RansburgThomas RathgenPeter AmonAndreas HutterHermann Hellwagner

Terms and definitions for the SVC file format

14496Thomas RathgenPeter AmonAndreas Hutter

On the SVC file format

14497 Philippe de Cuetos on behalf of FNB French NB comment on FCD 21000-14

14498

Khaled MamouTitus ZahariaMarius PredaFrançoise Prêteux

FAMC with streaming support

14499

Johannes HilpertSascha DischHeiko PurnhagenWerner Oomen

Proposed MPEG Surround Level Enhancement

14500 Sylvain Devillers Use of MPEG URN for identifying profiles and levels

14501 Anisse Taleb Report on the Evaluation of MPEG-4 Enhanced Low Delay AAC on Speech Content

14502

Daniel OanceaPedro CarvalhoTeresa AndradeChristian TimmererHermann Hellwagner

Defect Report on ISO/IEC 21000-15

14503

Hélder CastroPedro CarvalhoTeresa AndradeChristian TimmererHermann Hellwagner

A DID model for Media Streaming MAF

14504 Heiko Purnhagen Proposed Updates for MPEG Surround Conformance

41

Andreas SchneiderFrans de BontKarsten LinzmeierRalph Sperschneider

14505 Eva RodríguezJaime Delgado

Contribution to MPEG-21 Reference Software: Validation Rules Checker for the REL MAM Profile

14506 Yuriy Reznik Summary of core experiments on fixed point IDCT/DCT

14507 Eva RodríguezJaime Delgado

Contribution to the current version of the Open Release MAF

14508Eva RodríguezJaime DelgadoVíctor Torres

Some issues on the generation and modification of Event Reports in the MPEG-21 Event Reporting

14509 Yuriy Reznik Cross-check of IDCT conformance tests

14510 Yuriy Reznik Proposal for adding ISO/IEC 23002-2 in RVC tool library

14511 Florian Schreiner Chun Hui Suen

Overview of ISO/IEC 23000-7 CD Open Release MAF (1-pager)

14512 Gary J. SullivanRegis Crinon

Proposed technical alternative to MPEG-2 Systems DCOR 1 text WG 11 N 8859

14513 Florian Schreiner Chun Hui Suen

Proposed text to ISO/IEC 23000-7 CD Open Release MAF

14514 Markus SchnellRalf Geiger Proposed FPDAM of AAC-ELD

14515Markus SchmidtRalf GeigerMarkus Schnell

Cross-check report on Proposed FT Core Experiment for AAC-ELD

14516

Ralf GeigerMarkus SchnellJürgen HerreKristofer Kjörling

Utilizing AAC-ELD for delayless mixing in frequency domain

14517

Markus SchnellJürgen HerreRalf GeigerMarkus SchmidtMarkus Multrus

Proposed Core Experiment on AAC-ELD

14518Markus SchmidtRalf GeigerMarkus Schnell

Additional information on quality of AAC-ELD

14519Catherine Colomes Pierrick Philippe David Virette

Listening test results on instantaneous block switching CE for AAC ELD

14520 Pierrick Philippe David Virette

Updated description for AAC ELD instantaneous block switching CE

42

14521

Saar De ZutterFrederik De KeukelaereGerrard DruryChristian TimmererXin Wang

Editors' input to ISO/IEC FCD 21000-8 Reference Software (2nd edition)

14522Juha Ojanperä miikka.vilermo@nokia.com Miikka Vilermo

On AAC LTP conformance

14523

A-Young ChoIk-Hwan ChoJun-Woo LeeWeon-Geun OhDong-Seok Jeong

New Visual Identifier for MPEG-7 VCE-6 Basic Condition

14524 Saar De Zutter Review of Core Experiment on query operation based on text description

14525Ying Chen Ye-Kui Wang Miska M. Hannuksela

Signaling of leading pictures in file format

14526 Ye-Kui Wang Miska M. Hannuksela On SVC file format

14527 Ye-Kui Wang Miska M. Hannuksela

Signaling of temporal layer switching points in SVC file format

14528 Ye-Kui Wang Miska M. Hannuksela Alternate group parameters in ISO file format Amd. 2

14529 David Singer MP4 file format considerations for high sample-rate audio

14530

Henney OhYang-Won JungHyo Jin KimChang-Heon LeeHong-Goo Kang

Cross-check report on proposed FT Core Experiment for AAC-ELD

14531 Arianne T. Hinds Fixed-Point IDCT Conformance Tests

14532 Gerrard Drury Contribution on URI assets and Requirements and Structure of URNs

14533 the 3D Consortium via SC 29 Secretariat

Liaison Statement from the 3D Consortium [SC 29 N 8334]

14534 ISO TC 46/SC 9/WG 7 via SC 29 Secretariat

Liaison Statement from ISO TC 46/SC 9/WG 7 [SC 29 N 8335]

14535 JSR-287 EG via SC 29 Secretariat Liaison Statement from JSR 287 Expert Group [SC 29 N 8336]

14536 Frans de BontWerner Oomen

Cor to 14496-3:2005 subpart 10, DST (lossless oversampled audio)

14537 Masanori SanoHideki Sumiyoshi

Test report of CE on Query expression

43

Nobuyuki Yagi

14538Masanori SanoHideki SumiyoshiNobuyuki Yagi

Test report of CE on specification of the request of the Output

14539Masanori SanoHideki SumiyoshiNobuyuki Yagi

Test report of CE on Query operation based on text description

14540

Seungkwon BeackJeongil SeoTaejin Leekyungok kang

Information on SAOC test items by ETRI

44

Annex D – Output documents

No. Source Title

8910 Convener List of Documents from the San Jose, USA Meeting

8911 Convener Resolutions of the San Jose, USA

8912 Convener List of AHGs Established at the 80th Meeting in San Jose, USA

8913 Convener Report of the 80th Meeting in San Jose, USA

8914 Convener Guidelines for Electronic Distribution of MPEG and WG 11 Documents

8915 Convener Press Release of the 80th Meeting in San Jose, USA

8916 Convener Meeting Notice of the 81st Meeting in Lausanne, Switzerland

8917 HoD Guide for WG 11 Meeting Hosts

8918 HoD MPEG 101

8919 Liaison Liaison statement to WG1

8920 Liaison Liaison Statement to IETF

8921 Liaison Liaison Statement to Khronos

8922 Liaison Liaison Statement to ISO TC184 SC4

8923 Liaison Liaison Statement to 3GPP

8924 Liaison Liaison Statement to W3C

8925 Liaison Liaison Statement to ITU-T FG/IPTV concerning M3W

8926 Liaison Liaison Statement to ITU-T FG IPTV

8927 Liaison Liaison Statement to SMPTE

8928 Liaison Liaison Statement to DVD Forum

8929 Liaison Liaison Statement to ETSI

8930 Liaison Liaison Statement to SMPTE re File Format

8931 Liaison Liaison Statement to DVB

8932 Liaison Liaison Statement to JCP

8933 Liaison Liaison Statement to CEA

8934 Liaison Liaison Statement to ATIS

8935 Liaison Liaison Statement to SMPTE re RVC

8936 Liaison Liaison Statement to 3D consortium

8937 Liaison Liaison Statement to FLOForum

8938 Liaison Liaison Statement to TC46/SC9/WG7

8939 Liaison Liaison Statement to AVS

8940 Liaison Response to National Bodies

8941 Liaison Liaison Statement to DVB

45

8942 Requirement MAFs Overview

8943 Requirement RVC Requirements

8944 Requirement FTV Model and Requirements

8945 Requirement Requirements on and Structure for Assignment of MPEG URNs

8946 Convenor AHG on Review of MPEG-21 DID

8947 Convenor AHG on FTV

8948 Video Disposition of Comments on ISO/IEC 14496-2:2004/PDAM4

8949 Video Text of ISO/IEC 14496-2:2004/FPDAM4 Simple Profile Level 6

8950 Video Text of ISO/IEC 14496-4:2004/DCOR4

8951 Video Text of ISO/IEC 14496-4:2004/Amd.1/DCOR2

8952 Video Disposition of Comments on ISO/IEC 14496-4:2004/PDAM28

8953 Video Text of ISO/IEC 14496-4:2004/FPDAM28 Visual Simple Profile Level 6 Conformance Testing

8954 Video Request for ISO/IEC 14496-4:2004/Amd.30

8955 Video Working Draft 1 of ISO/IEC 14496-4:2004/Amd.30 AVC Professional Profiles Conformance Testing

8956 Video Request for ISO/IEC 14496-4:2004/Amd.31

8957 Video Working Draft 1 of ISO/IEC 14496-4:2004/Amd.31 SVC Conformance Testing

8958 Video Request for ISO/IEC 14496-5:2001/Amd.18

8959 Video Working Draft 1 of ISO/IEC 14496-5:2001/Amd.18 Professional Profiles Reference Software

8960 Video Request for ISO/IEC 14496-5:2001/Amd.19

8961 Video Working Draft 1 of ISO/IEC 14496-4:2001/Amd.19 SVC Reference Software

8962 Video Study Text (version 3) of ISO/IEC 14496-10:2005/FPDAM3 Scalable Video Coding

8963 Video Joint Scalable Video Model (JSVM) 10

8964 Video JSVM 10 Software

8965 Video Draft SVC Verification Test Plan Version 3.0

8966 Video Working Draft 3 of ISO/IEC 14496-10:2005/Amd.4 Multiview Video Coding

8967 Video Joint Multiview Video Model (JMVM) 4

8968 Video JMVM 4 Software

8969 Video Text of ISO/IEC 15938-3:2002/Amd.2:2006/Cor.1 (Perceptual 3D Shape)

8970 Video MPEG-7 Visual XM Document version 30.0

8971 Video Description of Core Experiments for MPEG-7 New Visual Extensions

8972 Video Disposition of Comments on ISO/IEC 15938-6:2003/ Amd.1:2006/DCOR 1

8973 Video Text of ISO/IEC 15938-6:2003/Amd.1:2006/Cor.1 (Color Temperature)

8974 Video Disposition of Comments on ISO/IEC 15938-6:2003/FPDAM2

46

8975 Video Text of ISO/IEC 15938-6:2003/FDAM2 (Perceptual 3D Shape)

8976 Video Disposition of Comments on ISO/IEC 15938-7:2003/FPDAM3

8977 Video Text of ISO/IEC 15938-7:2003/FPDAM3 (Perceptual 3D Shape)

8978 Video Study Text of ISO/IEC 23000-3/PDAM1 Reference Software for Photo Player MAF

8979 Video WD 4 of ISO/IEC 23001-4

8980 Video Disposition of Comments on ISO/IEC 23002-1/PDAM1

8981 Video Text of ISO/IEC 23002-1/FPDAM1 Software for Integer IDCT Accuracy Testing

8982 Video Disposition of Comments on ISO/IEC CD 23002-2

8983 Video Text of ISO/IEC FCD 23002-2 Fixed-point Implementation of 8x8 IDCT and DCT

8984 Video WD 4 of ISO/IEC 23002-4

8985 Video Description of Core Experiments in RVC

8986 Video RVC Simulation Model (RSM) V4.0

8987 Video RVC Work Plan

8988 Video RVC Conformance Testing Working Draft 1.0

8989 Video Description of Exploration Experiments for Toolbox Extensions

8990 Convenor AHG on Maintenance of MPEG-4 Visual related Documents, Reference Software and Conformance

8991 Convenor AHG on Reconfigurable Video Coding

8992 Convenor AHG on MPEG-7 Visual and Photo Player MAF

8993 Convenor AHG on SVC Verification Test

8994 ISG Status of HDL submissions and commitments for MPEG-4 Part-9

8995 ISG Study of “ISO/IEC DTR 14496-9 3rd Edition Reference Hardware Description”

8996 Convenor AHG on MPEG-4 Part 9: Reference Hardware Description Phase 1 and 2.

8997 Convenor AHG for Video Annotation

8998 Systems Text of ISO/IEC 13818-1:2003/DCOR1.2 (AVC Referencing and PS Signaling)

8999 Systems DoC on ISO/IEC 14496-4/PDAM.23 Synthesised Texture Conformance

9000 Convenor Terms of Reference

9001 Convenor MPEG Standards

9002 Convenor Table of unpublished FDIS

9003 Convenor Work plan and time line

9004 Convenor Work item assignment

9005 Convenor MPEG Standard Editors

9006 Convenor Software assets

47

9007 Convenor Conformance assets

9008 Convenor Content assets

9009 Convenor URI assets

9010 Convenor Standards under development for which a call for patent statements is issued

9011 Convenor List of Organisations with which MPEG entertains liaisons

9012 Systems Text of ISO/IEC 14496-4/FPDAM.23 Synthesised Texture Conformance

9013 Systems DoC on ISO/IEC 14496-4/PDAM.24 File Format Conformance

9014 Systems Text of ISO/IEC 14496-4/FPDAM.24 File Format Conformance

9015 Systems DoC on ISO/IEC 14496-4/PDAM.25 LASeR V1 Conformance

9016 Systems Text of ISO/IEC 14496-4/FPDAM.25 LASeR V1 Conformance

9017 Systems Request for ISO/IEC 14496-4/Amd.26

9018 Systems Text of ISO/IEC 14496-4/PDAM.26 Open Font Format Conformance

9019 Systems DoC of ISO/IEC 14496-5/FPDAM12 File Format Reference Software

9020 Systems Text of ISO/IEC 14496-5/FDAM12 File Format Reference Software

9021 Systems Text of ISO/IEC 14496-11/COR.6 (AudioFXProto correction and Bitwrapper)

9022 Systems DoC on ISO/IEC 14496-12/FPDAM2 (Flute Hint Track)

9023 Systems Text of ISO/IEC 14496-12/FDAM2 (Flute Hint Track)

9024 Systems Text of ISO/IEC 14496-12/COR.3

9025 Systems TuC for ISO/IEC 14496-12 & 15444-12

9026 Systems Study Text of ISO/IEC 14496-15/PDAM2 (SVC File Format)

9027 Systems ISO/IEC 14496-20/DCOR2

9028 Systems DoC on ISO/IEC 14496-20/FPDAM1 (LASeR Extensions)

9029 Systems Text of ISO/IEC 14496-20/FDAM1 (LASeR Extensions)

9030 Systems Request for ISO/IEC 14496-20/Amd.2 (SVGT1.2 Support)

9031 Systems Text of ISO/IEC 14496-20/FPDAM2 (SVGT1.2 Support)

9032 Systems TuC for ISO/IEC 14496-20/Amd2

9033 Systems WD3.0 of ISO/IEC 14496-20 2nd Edition (1st Ed. + Cor + Amd.1)

9034 Systems IuC for LASeR

9035 Systems Request of ISO/IEC 21000-9/Amd.1

9036 Systems Text of ISO/IEC 21000-9/PDAM.1 Mime Type Registration

9037 Systems DoC of ISO/IEC 23000-4/FCD (Musical Slide Show MAF)

9038 Systems Text of ISO/IEC 23000-4/FDIS (Musical Slide Show MAF)

9039 Systems Workplan for Musical Slide Show MAF Conformance and Ref. Software

9040 Systems WD1.0 of ISO/IEC 23000-4/Amd.2 Protected Musical Slide Show

9041 Systems Text of ISO/IEC 23000-8/CD (Portable Video Player MAF)

48

9042 Systems DoC on ISO/IEC 23000-9/CD (MAF for DMB)

9043 Systems Text of ISO/IEC 23000-9/FCD (MAF for DMB)

9044 Systems TuC on MAF for DMB

9045 Systems Request for ISO/IEC 23000-10

9046 Systems WD1.0 on ISO/IEC 23000-10 (Video Surveillance MAF)

9047 Systems Study Text of ISO/IEC 23001-1/FPDAM2 (Prefixes and of wild cards extensions)

9048 Systems DoC on ISO/IEC 23001/DCOR2

9049 Systems Text of ISO/IEC 23001/COR2

9050 Systems DoC on ISO/IEC 23001-2/FCD (Fragment Request Unit)

9051 Systems Text of ISO/IEC 23001-2/FDIS (Fragment Request Unit)

9052 Systems Text of ISO/IEC 23001-3/FCD (IPMP XML Messages)

9053 Systems Text of ISO/IEC 23004-5/FDIS Component Download

9054 Systems Text of ISO/IEC 23004-6/FDIS Fault Management

9055 Systems Text of ISO/IEC 23004-7/FDIS System Integrity Management

9056 Systems WD2.0 of ISO/IEC 23004-8 Reference Software and Conformance

9057 Systems M3W Reference Software and Conformance Plan

9058 Systems DoC on ISO/IEC 29116-1/CD Media Streaming MAF Protocol

9059 Systems Text of ISO/IEC 29116-1/FCD Media Streaming MAF Protocol

9060 Systems A project to exploit MPEG standards in tune with industry practices and needs

9061 Convenor Ad Hoc Group on Scene Representation

9062 Convenor Ad Hoc Group on MPEG File Formats

9063 Convenor Ad Hoc Group on MAF Under Development in Systems

9064 Audio DoC on ISO/IEC 11172-5:199x/DCOR 1

9065 Audio ISO/IEC 11172-5:199x/Cor. 1

9066 Audio DoC ISO/IEC 13818-7:2006/FPDAM 1

9067 Audio ISO/IEC 13818-7:2006/FDAM 1, Transport of MPEG Surround data in AAC

9068 Audio ISO/IEC 14496-3:2005/DCOR 6, DST and MP3on4

9069 Audio ISO/IEC 14496-3:2005/DCOR 7, SLS

9070 Audio DoC on ISO/IEC 14496-3/PDAM 8

9071 Audio ISO/IEC 14496-3/FPDAM 8, MP4FF Box for Original Audio File Information

9072 Audio DoC on ISO/IEC 14496-3:2005/PDAM 9 Request for Amendment.

9073 Audio DoC on ISO/IEC 14496-3:2005/PDAM 9

9074 Audio ISO/IEC 14496-3:2005/FPDAM 9, AAC-ELD

9075 Audio WD on MPEG-4 Audio Fourth Edition

9076 Audio DoC on ISO/IEC 14496-4:2004/FPDAM 14

49

9077 Audio ISO/IEC 14496-4:2004/FDAM 14, BSAC Extensions Conformance

9078 Audio DoC ISO/IEC 14496-4:2004/FPDAM 18

9079 Audio ISO/IEC 14496-4:2004/FDAM 18, MPEG-1 and -2 on MPEG-4 Conformance

9080 Audio DoC ISO/IEC 14496-4:2004/FPDAM 19

9081 Audio ISO/IEC 14496-4:2004/FDAM 19, ALS Conformance

9082 Audio Study on ISO/IEC 14496-4:2004/FPDAM 20, SLS Conformance

9083 Audio Status of MPEG-4 Audio Conformance

9084 Audio Status of MPEG-4 SLS Conformance

9085 Audio ISO/IEC 14496-5:2001/AMD 10:2007/DCOR 1, BSAC and SLS

9086 Audio Request for Amendment, MPEG-1/2 on MPEG-4 Ref. Software

9087 Audio ISO/IEC 14496-5:2001/AMD XX, MPEG-1/2 on MPEG-4 Ref. Software

9088 Audio DoC ISO/IEC FCD 14496-23

9089 Audio ISO/IEC FDIS 14496-23:200x, Symbolic Music Representation

9090 Audio DoC ISO/IEC 23003-1:2007/PDAM 1

9091 Audio ISO/IEC 23003-1:2007/FPDAM 1, MPEG Surround Conformance

9092 Audio DoC ISO/IEC 23003-1:2007/PDAM 2

9093 Audio ISO/IEC 23003-1:2007/FPDAM 2, MPEG Surround Reference Software

9094 Audio Defect Report of ISO/IEC 23003-1:2007

9095 Audio Framework for Exploration of Speech and Audio Coding

9096 Audio Workplan for Exploration of Speech and Audio Coding

9097 Convenor AHG on Audio Standards Maintenance

9098 Convenor AHG on SAOC CfP, AAC-ELD and Speech and Audio Exploration

9099 Audio Final Spatial Audio Object Coding Evaluation Procedures and Criterion

9100 MDS ISO/IEC FPDAM/1 15938-5 Improvements to Geographic Descriptor

9101 MDS ISO/IEC FPDAM/1 15938-7 Improvements to Geographic Descriptor Conformance

9102 MDS Schema Files for MPEG-7

9103 MDS ISO/IEC 15938-12 CD MPEG-7 Query Format

9104 MDS Technologies Under Consideration for MPEG-7 Query Format

9105 MDS DoC of ISO/IEC 21000-4 FPDAM/1 IPMP Components Base Profile

9106 MDS Text of ISO/IEC 21000-4 FDAM/1 IPMP Components Base Profile

9107 MDS DoC of ISO/IEC 21000-5 PDAM/3 ORC (Open Release Content) Profile

9108 MDS ISO/IEC 21000-5 FPDAM/3 ORC (Open Release Content) Profile

9109 MDS Interoperability between MPEG-21 REL DAC Profile and other Rights Information Standards

9110 MDS REL/RDD Reference Software Development Plan v.6

50

9111 MDS Disposition of Comments on ISO/IEC 21000-7:2004/DCOR 1

9112 MDS Text of ISO/IEC 21000-7:2004/COR 1 MPEG-21 Digital Item Adaptation

9113 MDS Text of ISO/IEC 21000-7 FDIS Second edition

9114 MDS Preliminary DoC of preliminary comments of ISO/IEC 21000-8 FCD Reference Software

9115 MDS Study text of ISO/IEC 21000-8 FCD Reference Software

9116 MDS Doc of ISO/IEC 21000-14 Conformance

9117 MDS Text of ISO/IEC FDIS 21000-14 Conformance

9118 MDS ISO/IEC 21000-15:2006/DCOR1 MPEG-21 Event Reporting

9119 MDS DoC of ISO/IEC 21000-18/PDAM 1

9120 MDS ISO/IEC 21000-18/FPDAM/1 Simple fragmentation rule

9121 MDS DoC of ISO/IEC 23000-2 FCD Music Player Application Format 2nd Edition

9122 MDS Text of ISO/IEC 23000-2 FDIS Music Player Application Format 2nd Edition

9123 MDS DoC on ISO/IEC CD 23000-5 Media Streaming Player

9124 MDS ISO/IEC FCD 23000-5 Media Streaming Player

9125 MDS DoC of ISO/IEC 23000-7 CD Open release MAF

9126 MDS ISO/IEC 23000-7 FCD Open release MAF

9127 MDS Text of ISO/IEC 23001-5 FDIS Bitstream Syntax Description Language

9128 Convenor AHG on MPEG-7 Query Format

9129 MDS DoC ISO/IEC PDAM/1 15938-5 Improvements to Geographic Descriptor

9130 MDS DoC ISO/IEC PDAM/1 15938-7 Improvements to Geographic Descriptor Conformance

9131 Requirements MPEG Profiles and Levels URIs

9132 3DGC Text of ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

9133 3DGC Text of ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

9134 3DGC Text of ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)

9135 3DGC Text of ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)

9136 3DGC WD 2.0 of ISO/IEC 14496-16:2006/AMD2 (Frame-based Animated Mesh Compression)

9137 3DGC WD 1.0 of ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

9138 3DGC 3D Graphics Core Experiments Description

9139 3DGC 3D Graphics Compression FAQ 19.0

9140 3DGC Text of ISO/IEC 14496-21:2006/COR1

9141 3DGC Request for Subdivision of ISO/IEC 14496: Part 25 - 3D Graphics Compression Model

9142 3DGC WD 1.0 for ISO/IEC 14496-25

51

9143 Convenor AHG on 3DG documents, experiments and software maintenance

9144 Systems TuC for IPMP XML Messages      

9145 Convenor Project Editors for ISO/IEC Certificate of Appreciation

9146 3DGC DoC on ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

9147 3DGC DoC on ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

9148 3DGC DoC on ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)

9149 3DGC Doc on ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)

9150 3DGC Request for ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

9151 MDS Request of subdivision for MPEG-7 Query Format

9152 Systems Disposition of Comments NWIP, Information technology -- Supplemental media technologies

9153 Systems Elements for a solution for storage of MPEG-2 TS in the MPEG-4 File Format

52

Annex E – Requirements report

Source: Fernando Pereira (Instituto Superior Técnico, Lisboa-Portugal

Note: Requirements agenda for the San Jose MPEG meeting is annexed at the end of this report.

16 Requirements documents approved at this meeting8942 MAFs Overview8943 RVC Requirements8944 FTV Model and Requirements8945 Requirements on and Structure for Assigning MPEG URNs9131 MPEG Profiles and Levels URIs

17 MPEG Structure

17.1 MPEG URNs and URIs (joint with MDS)14500, Sylvain Devillers, Use of MPEG URN for identifying profiles and levels MPEG video and audio coding formats are used by a large number of standards developed by other bodies such as DBV and 3GPP. Such standards may normatively reference a video or audio coding format, but in some cases reference a given profile and level of such format. This contribution proposed that, to promote the adoption of WG11 standards, it is the responsibility and interest of the WG11 to define, publish and maintain a list of unique identifiers for profiles and levels of MPEG coding formats. Following, this contribution, it was decided to create a document (N9131) with MPEG profiles and levels URIs. This document will include unique URIs for all MPEG profiles and levels. All the MPEG subgroups are kindly asked to review this document, especially in the parts regarding their own profiles and levels.

14532, Gerrard Drury, Contribution on URI assets and Requirements and Structure of URNsThe use of Uniform Resource Identifiers (URIs) within MPEG standards has become more prevalent, particularly with the increased use of XML in MPEG standards. Because there was no global standard structure for the URNs being used in MPEG standards, a document was created at the last meeting (N8785) including motivation, objectives and process to define URNs, requirements on URNs, definition of required URNs structure, and URN examples. This contribution proposed some corrections and improvements to the document issued at last meeting that have been approved. Following this approval a revised version of the Requirements on and Structure for Assigning MPEG URNs document (8945) has been issued.

18 MPEG-4

18.1 Metadata in AVC (joint with Video & JVT)Some contributions regarding metadata in AVC were submitted to JVT at this meeting. During a joint meeting with Video and JVT, it was concluded that AVC metadata shall be based on MPEG-7 tools and thus the issue is to be addressed in MPEG; coding related metadata may need especial consideration when its purpose target coding efficiency. The next steps for this activity may include:

1. Identification of requirements at various levels2. Understanding if new MPEG-7 tools are needed to address requirements

53

3. Understanding if new MPEG-7 profile is needed

18.2 3D Compression Profiling (joint with 3DGC)14467, Patrick Gioia, Olivier Aubault, Proposal for 3D Compression ProfileThis contribution proposes profiles in the 3D area to address Google-earth like applications, in real-time, with adaptive navigation. Following discussions at last meeting, it was agreed that the full picture in the 3 graphics related profiling dimensions has to be kept in mind to cover well the profiling space. Following the discussions, profiles in the 3 graphics related profiling dimensions will be defined, notably:

Basic AFX in Scene Graph Basic AFX in Graphics Multires in 3D Compression (with 8 object types and 2 levels)

18.3 Laser (joint with Systems)14373, Jean-Claude Dufourd, LASeR profiles adjustmentLaser version 1 includes currently the Mini and Full profiles. Following this contribution and discussions at last meeting, it was decided:

1. To correct MINI profile to make it useful and hierarchical to Core2. To remove FULL profile because useless and illy defined3. To define CORE profile (hierarchical to MINI)4. To start studying possible MAIN profile (hierarchical to CORE)

19 MPEG-7

19.1 MP7QF 14420, Hee-Cheol Seo, Miran Choi, Hyunki Kim, Myung-Gil Jang, Soojong Lim, Jeong Heo, Kyoungro Yoon, CE Report for Query Expression of MPEG-7 Query Format 14421, Hee-Cheol Seo, Miran Choi, Hyunki Kim, Myung-Gil Jang, Soojong Lim, Jeong Heo, Kyoungro Yoon, Revision of Proposed Input Query Format for MPEG-7 Query Format14449, Doeller, Gruhne, Wolf, MP7QF CE Test ReportThese contributions have been addressed by the MDS subgroup since they include technical inputs related to an activity managed by MDS.

20 MPEG-21

20.1 Digital Item Declaration14475, Giovanni Cordara (on behalf of the ITNB), Italian NB proposal to revisit MPEG-21 DIDThis contribution states that “Italy believes that it would be beneficial to revisit the MPEG-21 DID requirements on the basis of the experience gathered with ISO/IEC 21000-1 and propose a New Project that aims at a new standard with the functionalities derived from the revisiting of the MPEG-21 DID requirements and with the constraint that no IP contained in patents whose rights are currently valid be required to implement the new standard or, if such IPR exists, it is licensed by its holder royalty free.”Following this contribution, a BoG was established to:

1. Identify possible DID deficiencies and possible solutions

54

2. Revisit DID requirements3. Assess the feasibility of reaching the target proposed by ITNB

To continue the work from the BoG, an AHG has been established (N8946) with the following mandates:

1. Investigate whether the current DID (ISO/IEC 21000-2) requirements fit with today’s industry, and if not review the requirements.

2. Collect information on how DID is currently used.3. Identify current deficiencies with DID and propose ways to address these deficiencies.4. Investigate feasibility of producing royalty-free DID.

21 MPEG-A

21.1 Professional Archival MAF14430, Tilman Liebchen, Comments on Professional Archival MAF Requirements14411, Noboru Harada, Takehiro Moriya and Yutaka Kamamoto, Proposed text to WD of Professional Archival MAFAlthough this MAF is already under development by MDS, it was discussed in a joint meeting with MDS to review the requirements and check the industry support. It was confirmed that there is currently no significant industry support for the current set of requirements. The experts involved in this MAF committed to bring at the next meeting further requirements contributions and evidence of more industry support.

21.2 Surveillance MAF14352, James Orwell, Contribution to the Basic Video Surveillance MAFThis contribution was not presented because the author was not available.

14486, Houari Sabirrin, Jeongyeon Lim, Munchurl Kim, A Proposal for Basic Video Surveillance Application FormatFollowing this contributions and discussions in a BoG, it was decided to promote to ‘under development’ a rather simple MAF to package surveillance video content, mainly including the following tools: AVC file format, AVC video (Baseline profile) and some MPEG-7 metadata. It is recognized that this simple MAF may be important to penetrate in a rather new application domain for MPEG: surveillance. Since there is support to create in the future a more complete MAF for surveillance applications, e.g. including audio, there is still a surveillance related MAF under consideration, now renamed ‘Advanced surveillance’.

21.3 Protected Musical Slide Show MAF14477, Hendry, Houari Sabirin, Munchurl Kim, Updated Proposal for Protected Musical Slide Show MAF with IPMPFollowing evidence of need and industry support, this MAF was promoted to ‘under development’. This MAF adds protection capabilities to the Musical Slide Show already under development by the Systems subgroup. It was agreed that the technical solution for the additional protection capabilities will be siilar to the solution used for the Music Player MAF.

55

21.4 Protected Photo Player MAF14478, Hendry, Houari Sabirin, Munchurl Kim, Updated Proposal for Protected Photo Player MAF with IPMPFollowing the discussion, this MAF stays ‘under consideration’ since it needs clear industry support and also to address technical issues raised by the MPEG-7 Visual BoG.

21.5 Stereoscopic MAF14423, Kwangcheol Choi, Sung-Moon Chun, Jaedo Kwak, Seungheon Yang, Ji-Sang Yoo, Si-Hun Sung, Seong-Cheol, Han, Requirements for Stereoscopic MAF14424, Jaedo Kwak, Si-Hun Sung, Sung-Moon Chun, JinWoong Kim, Namho Hur, Whitepaper of Stereoscopic ProjectFollowing evidence of market need and industry support, this MAF was promoted to ‘under consideration’. Further contributions are expected at the next meeting (notably in terms of technical solutions) in order further progress may be made.

21.6 Cross-Media Interactive Presentation MAF14367, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Kia Ng (University of Leeds), Proposal for a MAF on Cross-Media Interactive Presentation: Application Scenarios14368, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Proposal for a MAF on Cross-Media Interactive Presentation: Requirements14369, Davide Rogai, Pierfrancesco Bellini, Paolo Nesi, Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFsAlthough this new MAF proposal was discussed in three sessions, it was not possible to identify the scope and main functional target of this MAF. There was also no clear industry support for this MAF. Further progress on this MAF will require solving these two issues.

21.7 Summary on MAFsThe global MAF situation after the San Jose MPEG meeting is summarized in the MAFs Overview document (N8942) as follows:

1. MAFs Finalizeda. Music Player MAF (including protection) b. Photo Player MAF

2. MAFs Under Developmenta. Photo Player MAF (under Video) b. Musical Slide Show MAF, including protection (under Systems) c. Media Streaming MAF (under MDS) d. Professional Archival MAF (under MDS) e. Open Release MAF (under MDS) f. Portable Video Player MAF (under Systems) g. MAF for Digital Multimedia Broadcasting (under Systems)h. Video Surveillance MAF (under Systems)

3. MAFs Under Considerationa. Advanced Surveillance MAFb. Protected Photo Player MAF

56

c. Digital Video/Cinema MAFd. Stereoscopic MAF

22 MPEG-B and MPEG-C

22.1 RVC (joint with Video/ISG)14511, AVS Working Group, Liaison Statement to MPEG on RVCFollowing this contribution from AVS, MPEG states that the RVC project is about developing A full collection of MPEG individual coding tools organized in the MPEG video tool library

and A generic framework that can be used to make an implementation of any MPEG video coding

standard and additionally is capable of supporting the implementation of video coding standards from other organizations with which a collaboration can be established.

As part of this project, an identification mechanism will be developed whereby MPEG video coding tools will be identified by MPEG and video coding tools from other organizations can be identified via a registration authority.

23 Explorations

23.1 Freeviewpoint Television (FTV)14417, Masayuki Tanimoto, Toshiaki Fujii, Hideaki Kimata, Shigeyuki Sakazawa, Proposal on Requirements for FTV14533, Liaison from 3D Consortium Based on these contributions, it was agreed that FTV is an important application domain which MPEG has been trying to address since a long time. Following recent inputs, there is a need to revisit the way MPEG may address this application domain using existing MPEG standards and very likely adding new standards. In conclusion, FTV is currently an MPEG activity, targetting at this stage to

1. Identify an FTV architecture and model2. Identify for which architectural modules normative technology should be specified, e.g.

FTV data format, decoding, rendering3. Identify the requirements for each normative modules from the visual, audio and systems

perspectivesAfter the issues above are clarified, the FTV roadmap will be defined, notably the relation with JVT activities. A response to the 3D Consortium has been prepared describing the activities MPEG is currently developing in this area. An AHG (N8947) has been created with the following mandates:1. To refine the FTV architecture.2. To refine the identification and definition of normative elements in the FTV architecture 3. To refine the FTV requirements

23.2 Full Colour Video Coding14452, Tokumichi Murakami, Kohtaro Asai, Yoshihisa Yamada, Requirement of Full-Color Video Coding for Consumer Applications

57

This contribution proposed requirements for a possible “full color” video coding standard adapted for consumer applications. The discussion confirmed these requirements are still drafty and thus further contributions are welcome at the next meeting.

23.3 IPTV Requirements

23.4 This activity reviewed and answered the liaison contributions on IPTV Requirements from ATIS/IIF IPTV, CEA and the ITU-T IPTV Focus Group. It was agreed there is a need to continue identifying the relevant requirements for MPEG from the inputs provided and checking the coverage of relevant requirements by existing MAFs, notably the Media Streaming MAF.

23.5 Dual-Track Licensing Approach14360, USNB Contribution: Issues relating to expiring patentsThe USNB contribution states that “if it is technically possible to develop a standard which does this (royalty free), the USNB prefers that it be done in WG 11 where there is expertise in doing it well, and where such a putative standard could be made a 'family member' with other MPEG standards (with an upgrade path, for example, or related technical ‘roots’ etc.)” and “the 'terms of engagement' of a study on developing a process for royalty-free standards, and the results and follow-on for such work, should be made more clear before more discussion is held at WG 11.”After discussion and based on past experience on the dual-track approach, it was decided that no further progress is possible in this activity until sufficient commitment is made available.

24 80th MPEG (San Jose) Agenda Requirements

25 Room: Oak

TIME TOPIC ROOM

Monday

  9:00-end Opening Plenary Meeting  

11:00-12:00

DID

14475, Giovanni Cordara (on behalf of the ITNB), Italian NB proposal to revisit MPEG-21 DID

Reqs

Lunch  

58

14:30-16:30

NEW MAF PROPOSALs

Stereoscopic MAF

14423, Kwangcheol Choi, Sung-Moon Chun, Jaedo Kwak, Seungheon Yang, Ji-Sang Yoo, Si-Hun Sung, Seong-Cheol, Han, Requirements for Stereoscopic MAF

14424, Jaedo Kwak, Si-Hun Sung, Sung-Moon Chun, JinWoong Kim, Namho Hur, Whitepaper of Stereoscopic Project

MAF on Cross-Media Interactive Presentation

14367, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Kia Ng (University of Leeds), Proposal for a MAF on Cross-Media Interactive Presentation: Application Scenarios

14368, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Proposal for a MAF on Cross-Media Interactive Presentation: Requirements

14369, Davide Rogai, Pierfrancesco Bellini, Paolo Nesi, Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFs

Reqs

16:30-18:00 BoGs -

18:00-20:00 HoDs Meeting HoD

Tuesday

9:00-11:00

Various (joint with MDS)URNs

14500, Sylvain Devillers, Use of MPEG URN for identifying profiles and levels

14532, Gerrard Drury, Contribution on URI assets and Requirements and Structure of URNs

MP7QF

14420, Hee-Cheol Seo, Miran Choi, Hyunki Kim, Myung-Gil Jang, Soojong Lim, Jeong Heo, Kyoungro Yoon, CE Report for Query Expression of MPEG-7 Query Format

14421, Hee-Cheol Seo, Miran Choi, Hyunki Kim, Myung-Gil Jang, Soojong Lim, Jeong Heo, Kyoungro Yoon, Revision of Proposed Input Query Format for MPEG-7 Query Format

14449, Doeller, Gruhne, Wolf, MP7QF CE Test Report

Reqs

12:00-13:00RVC and AVS (joint with ISG & Video)

14511, AVS Working Group, Liaison Statement to MPEG on RVCReqs

13:00-14:00 Lunch 14:00-18:00

MPEG-A (joint with MDS, Systems, Audio and Video)

MAFs UNDER CONSIDERATION

Surveillance MAF

14352, James Orwell, Contribution to the Basic Video Surveillance MAF

Reqs

59

14486, Houari Sabirrin, Jeongyeon Lim, Munchurl Kim, A Proposal for Basic Video Surveillance Application Format

Protected Musical Slide Show MAF

14477, Hendry, Houari Sabirin, Munchurl Kim, Updated Proposal for Protected Musical Slide Show MAF with IPMP

Protected Photo Player MAF

14478, Hendry, Houari Sabirin, Munchurl Kim, Updated Proposal for Protected Photo Player MAF with IPMP

MAFs UNDER DEVELOPMENT

Professional Archival MAF

14430, Tilman Liebchen, Comments on Professional Archival MAF Requirements

14411, Noboru Harada, Takehiro Moriya and Yutaka Kamamoto, Proposed text to WD of Professional Archival MAF

NEW MAF PROPOSALs

Stereoscopic MAF

14423, Kwangcheol Choi, Sung-Moon Chun, Jaedo Kwak, Seungheon Yang, Ji-Sang Yoo, Si-Hun Sung, Seong-Cheol, Han, Requirements for Stereoscopic MAF

14424, Jaedo Kwak, Si-Hun Sung, Sung-Moon Chun, JinWoong Kim, Namho Hur, Whitepaper of Stereoscopic Project

MAF on Cross-Media Interactive Presentation

14367, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Kia Ng (University of Leeds), Proposal for a MAF on Cross-Media Interactive Presentation: Application Scenarios

14368, Paolo Nesi, Pierfrancesco Bellini, Davide Rogai, Proposal for a MAF on Cross-Media Interactive Presentation: Requirements

14369, Davide Rogai, Pierfrancesco Bellini, Paolo Nesi, Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFs

18:00-19:00 Liaison Meeting19:00-end Chairs Meeting

Wednesday

  09:00-end plenary Plenary Meeting

-

12:00-12:30Profiling (joint with 3DGC)

14467, Patrick Gioia, Olivier Aubault, Proposal for 3D Compression Profile3DGC

Lunch14:00-15:30

Various (joint with Video, JVT)

14417, Masayuki Tanimoto, Toshiaki Fujii, Hideaki Kimata, Shigeyuki Sakazawa, Proposal on Requirements for FTV

JVT

60

14533, Liaison from 3D Consortium

14452, Tokumichi Murakami, Kohtaro Asai, Yoshihisa Yamada, Requirement of Full-Color Video Coding for Consumer Applications

14360, USNB Contribution: Issues relating to expiring patents

15:30-16:00

Carriage of MPEG-7 metadata in AVC (joint with Video, JVT) JVT

BoGsSocial Event

Thursday

9:00-9:30LASeR (joint with Systems)

14373, Jean-Claude Dufourd, LASeR profiles adjustmentRqs

9:30-12:00 Joint JPEG – MPEG on JPSearch Reqs

Lunch

14:00-15:00 Feedback from IPTV Requirements Reqs

15:00-16:00 Feedback from DID BoG (joint Reqs & MDS) Reqs

16:00-17:00

MAFs BoG Feedback

Surveillance MAF Protected Photo Player MAFCross-media Interactive Presentation MAF

Reqs

17:00-18:00 Reviewing FTV Requirements Doc Reqs

18:00-end Chairs Meeting

Friday

- Concluding MPEG-4- Reqs

- Concluding MPEG-7- MDS

9:00-9:15

Concluding MPEG-21

Response to Italian NB on new DID technologies – Giovani

AHG on New DID Technologies - Gerrard

Reqs

9:15-9:45Concluding MPEG-A

MAFs Overview - Florian Reqs

9:45-10:00RVC (MPEG-B & MPEG-C)

RVC Requirements - Euee Reqs

61

10:00-11:00

ExplorationsIPTV related Liaisons – Xin, Anthony

Response to US NB on royalty free standards

Revised Doc with URNs structure – Christian

FTV Model and Requirements – Tanimoto-san

AHG on FTV

Response to Liaison from 3D Consortium

Reqs

-12:00 -14:00 Lunch

14:00-end plenary Plenary Meeting

62

Annex F – Systems report

Source: Systems Chair and Break-out group ChairsContributors: David Singer (Apple), Young-Kwon Lim (Net&TV), Jean Gelissen (Philips)

1 OverviewThe main outputs of the meeting from the Systems Sub-group perspective are:

No. Title TBPX 13818-1 Systems8998 Text of ISO/IEC 13818-1:2003/DCOR1.2 (AVC Referencing and PS Signaling) NoX 14496-4 Conformance testing8999 DoC on ISO/IEC 14496-4/PDAM 23 Synthesised Texture Conformance No9012 Text of ISO/IEC 14496-4/FPDAM 23 Synthesised Texture Conformance No9013 DoC on ISO/IEC 14496-4/PDAM 24 File Format Conformance No9014 Text of ISO/IEC 14496-4/FPDAM 24 File Format Conformance No9015 DoC on ISO/IEC 14496-4/PDAM 25 LASeR V1 Conformance No9016 Text of ISO/IEC 14496-4/FPDAM 25 LASeR V1 Conformance No9017 Request for ISO/IEC 14496-4/Amd.26 No9018 Text of ISO/IEC 14496-4/PDAM 26 Open Font Format Conformance NoX 14496-5 Reference Software9019 DoC of ISO/IEC 14496-5/FPDAM 12 File Format Reference Software No9020 Text of ISO/IEC 14496-5/FDAM 12 File Format Reference Software NoX 14496-11 Scene Description and Application Engine9021 Text of ISO/IEC 14496-11/COR.6 (AudioFXProto correction and Bitwrapper) NoX 14496-12 ISO Base Media File Format9022 DoC on ISO/IEC 14496-12/FPDAM 2 (Flute Hint Track) No9023 Text of ISO/IEC 14496-12/FDAM 2 (Flute Hint Track) No9024 Text of ISO/IEC 14496-12/COR.3 No9025 TuC for ISO/IEC 14496-12 & 15444-12 NoX 14496-15 AVC File Format9026 Study Text of ISO/IEC 14496-15/PDAM 2 (SVC File Format) NoX 14496-20 Lightweight Application Scene Representation9027 ISO/IEC 14496-20/DCOR 2 No9028 DoC on ISO/IEC 14496-20/FPDAM 1 (LASeR Extensions) No9029 Text of ISO/IEC 14496-20/FDAM 1 (LASeR Extensions) No9030 Request for ISO/IEC 14496-20/Amd.2 (SVGT1.2 Support) No9031 Text of ISO/IEC 14496-20/FPDAM 2 (SVGT1.2 Support) Yes9032 TuC for ISO/IEC 14496-20/Amd.2 No9033 WD3.0 of ISO/IEC 14496-20 2nd Edition (1st Ed. + Cor + Amd.1) Yes9034 IuC for LASeR NoX 21000-9 File Format9035 Request of ISO/IEC 21000-9/Amd.1 No9036 Text of ISO/IEC 21000-9/PDAM 1 Mime Type Registration NoX 23000-4 Musical Slide Show MAF9037 DoC of ISO/IEC 23000-4/FCD (Musical Slide Show MAF) No9038 Text of ISO/IEC 23000-4/FDIS (Musical Slide Show MAF) No9039 Workplan for Musical Slide Show MAF Conformance and Ref. Software No9040 WD1.0 of ISO/IEC 23000-4/Amd.2 Protected Musical Slide Show NoX 23000-8 Portable Video Player9041 Text of ISO/IEC 23000-8/CD (Portable Video Player MAF) NoX 23000-9 Digital Multimedia Broadcasting Application Format9042 DoC on ISO/IEC 23000-9/CD (MAF for DMB) No9043 Text of ISO/IEC 23000-9/FCD (MAF for DMB) No9044 TuC on MAF for DMB No

63

X 23000-10 Video Surveillance MAF9045 Request for ISO/IEC 23000-10 No9046 WD1.0 on ISO/IEC 23000-10 (Video Surveillance MAF) NoX 23001-1 Binary MPEG Format for XML9047 Study Text of ISO/IEC 23001-1/FPDAM2 (Prefixes and of wild cards extensions) No9048 DoC on ISO/IEC 23001/DCOR2 No9049 Text of ISO/IEC 23001/COR2 NoX 23001-2 Fragment Request Unit9050 DoC on ISO/IEC 23001-2/FCD (Fragment Request Unit) No9051 Text of ISO/IEC 23001-2/FDIS (Fragment Request Unit) NoX 23001-3 IPMP XML Messages9052 Text of ISO/IEC 23001-3/FCD (IPMP XML Messages) No9144 TuC for IPMP XML Messages NoX 23004-5 Component Download9053 Text of ISO/IEC 23004-5/FDIS Component Download NoX 23004-6 Fault Management9054 Text of ISO/IEC 23004-6/FDIS Fault Management NoX 23004-7 Systems Integrity Management9055 Text of ISO/IEC 23004-7/FDIS System Integrity Management NoX 23004-8 Reference Software9056 WD2.0 of ISO/IEC 23004-8 Reference Software and Conformance No9057 M3W Reference Software and Conformance Plan NoX 29116-1 Media Streaming MAF Protocol 9058 DoC on ISO/IEC 29116-1/CD Media Streaming MAF Protocol No9059 Text of ISO/IEC 29116-1/FCD Media Streaming MAF Protocol NoX Exploration9060 A project to exploit MPEG standards in tune with industry practices and needs No

64

26 General issues

26.1 GeneralThe meeting report from Hangzou has been approved.The following demonstrations have been made:

None.

26.2 List of standards under developmentPr Pt Edit. Project Description CfP WD CD FCD FDIS2 1 2000 Cor.1 Reference to AVC

Specification07/04 07/10

4 4 2004 Amd.22 Audio BIFS v3 conformance

06/04 06/07 07/01 07/07

4 4 2004 Amd.23 Synthesized Texture conformance

06/07 07/01 07/04 07/10

4 4 2004 Amd.24 File Format Conformance 06/04 06/10 07/04 07/104 4 2007 Amd.25 LASeR V1 Conformance 06/04 06/10 07/04 07/104 4 2007 Amd.26 Open Font Format

Conformance07/04 07/10 08/01 08/07

4 4 2007 Amd.27 LASeR Amd.1 Conformance

06/10 07/07 07/10 08/04

4 5 2007 Amd.14 Open Font Format Ref. Soft 07/07 07/10 08/01 08/044 5 2007 Amd.16 Symbolic Music Rep. Ref.

Soft06/10 07/01 07/07 08/01

4 5 2007 Amd.17 LASeR Ref. Soft 06/10 07/01 07/07 08/014 15 2005 Amd.2 SCV File Format Extensions 05/10 06/07 07/07 08/014 20 2004 Cor.2 Profile Removal 07/04 07/104 20 2004 Amd.2 SVGT1.2 Support 05/10 07/04 07/1021 9 200x Amd.1 MP21 Mime Type 07/04 07/07 07/10A 4 200x Amd.1 MSS MAF Conf. and Soft 07/07 07/10 08/01 08/04A 4 200x Amd.2 Protected Musical Slide

Show07/04 07/07 07/10 08/04

A 8 200x 1st Ed. Portable Video Player MAF 06/10 07/04 07/10 08/01A 9 200x 1st Ed. Digital Multi. Broadcasting

MAF06/10 07/01 07/04 07/10

A 10 200x 1st Ed. Video Surveillance MAF 07/04 07/07 07/10 08/04B 1 200x Amd.2 Exten. On encoding of wild

cards06/04 06/07 07/01 07/07

B 3 200x 1st Ed. IPMP XML Messages 06/10 07/04 07/10E 8 200x 1st Ed. Ref. Soft. and Conformance 07/01 07/07 07/10 08/01X 1 200x Media Streaming MAF

Protocols06/10 07/04 07/10

65

26.3 Standing Documents

Pr Pt Documents No. Meeting1 1 MPEG-1 White Paper – Multiplex Format N7675 05/07 Nice1 1 MPEG-1 White Paper – Terminal Architecture N7676 05/07 Nice1 1 MPEG-1 White Paper – Multiplexing and

SynchronizationN7677 05/07 Nice

2 1 MPEG-2 White Paper – Multiplex Format N7678 05/07 Nice2 1 MPEG-2 White Paper – Terminal Architecture N7679 05/07 Nice2 1 MPEG-2 White Paper – Multiplexing and

SynchronizationN7680 05/07 Nice

2 11 MPEG-2 White Paper – MPEG-2 IPMP N7503 05/07 Poznan4 1 MPEG-4 White Paper – MPEG-4 Systems N7504 05/07 Poznan4 1 MPEG-4 White Paper – Terminal Architecture N7610 05/10 Nice4 1 MPEG-4 White Paper – M4MuX N7921 06/01 Bangkok4 1 MPEG-4 White Paper – OCI N8148 06/04 Montreux4 6 MPEG-4 White Paper – DMIF N8149 06/04 Montreux4 11 MPEG-4 White Paper – BIFS N7608 05/10 Nice4 12 MPEG-4 White Paper – ISO File Format N8150 06/04 Montreux4 14 MPEG-4 White Paper – MP4 File Format N7923 06/01 Bangkok4 15 MPEG-4 White Paper – AVC FF N7924 06/01 Bangkok4 13 White Paper on MPEG-4 IPMP N7505 05/07 Poznan4 13 MPEG IPMP Extensions Overview N6338 04/03 München4 17 White Paper on Streaming Text N7515 05/07 Poznan4 18 White Paper on Font Compression and Streaming N7508 05/07 Poznan4 20 Presentation Material on LASER N6969 05/01 Hong-

Kong4 20 White Paper on LASeR N7507 05/07 Poznan4 22 White Paper on Open Font Format N7519 05/07 Poznan7 1 MPEG-7 White Paper - MPEG-7 Systems N7509 05/07 Poznan7 1 MPEG-7 White Paper – Terminal Architecture N8151 06/04 Montreux21 9 MPEG-21 White Paper – MPEG-21 File Format N7925 06/01 BangkokB X MPEG-B White Paper – BinXML N7922 06/01 BangkokE X MPEG Multimedia Middleware Context and

ObjectivesN6335 04/03 München

E X 1rst M3W White paper N7510 05/07 PoznanE X 2nd M3W White Paper : Architecture N8152 06/04 MontreuxE X Tutorial on M3W N8153 06/04 MonreuxE X M3W White Paper : Multimedia Middleware

ArchitectureN8687 06/10 Hanzhou

E X M3W White Paper : Multimedia API N8688 06/10 HanzhouE X M3W White Paper : Component Model N8689 06/10 HanzhouE X M3W White Paper : Resource and Quality

ManagementN8690 06/10 Hanzhou

E X M3W White Paper : Component Download N8691 06/10 HanzhouE X M3W White Paper : Fault Management N8692 06/10 HanzhouE X M3W White Paper : System Integrity

ManagementN8693 06/10 Hanzhou

66

26.4 Mailing Lists Reminder

Topic Information Kindly Managed by

General Systems

List

Liste Reflector : gen-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/gen-sysmailto:gen-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/gen-sysList-Help: mailto:gen-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

BiM

Liste Reflector : mpeg7-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mpeg7-sysmailto:mpeg7-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mpeg7-sysList-Help: mailto:mpeg7-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

File Format

Liste Reflector : mp4-sys@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mp4-sysmailto:mp4-sys-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mp4-sysList-Help: mailto:mp4-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

LASeR

Liste Reflector : mpeg-laser@lists.uni-klu.ac.atList-Subscribe:

http://lists.uni-klu.ac.at/mailman/listinfo/mpeg-lasermailto:mpeg-laser-request@lists.uni-klu.ac.at?subject=subscribe

List-Archive: http://lists.uni-klu.ac.at/pipermail/mpeg-laserList-Help: mailto:mpeg-laser-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

MAF

Liste Reflector : maf-sys@lists.uni-klu.ac.atList-Subscribe:http://lists.uni-klu.ac.at/mailman/listinfo/maf-sysmailto:maf-sys-request@lists.uni-klu.ac.at?subject=subscribeList-Archive: http://lists.uni-klu.ac.at/pipermail/maf-sysList-Help: mailto:maf-sys-request@lists.uni-klu.ac.at?subject=help

University of Klagenfurt

26.5 FAQThe FAQ were updated as needed.

67

26.6 AOBNone.

27 MPEG-2 Systems (13818-1)

27.1 13818-1:2005 Amd.3 Carriage of SVC

27.1.1 Topics1. Transport of Scalable Video Coding

27.1.2 ContributionsM14305, M14349, M14329, M14382, M14512: Various input contributions related to the current DCOR. All of them requesting to find a solution that would not break backward compatibility. All of them where seriously reviewed during the meeting and proponents have been hardly working together to propose new text for this DCOR. This was successfully achieved and a new DCOR has been issued and submitted to ballot, replacing previous DCOR that will be abandoned.

Technical Work in Progress.

28 MPEG-4 Conformance (14496-4)

28.1 14496-4 Amd.22

28.1.1 Topics1. Audio BIFS Conformance

28.1.2 ContributionsNone.

Technical Work in Progress.

28.2 14496-4 Amd.23

28.2.1 Topics1. Synthesized Texture Conformance

28.2.2 ContributionsM14385: Summary of Voting on ISO/IEC 14496-4:2004/PDAM 23. No comment. Text of FPDAM produced.

Technical Work in Progress.

28.3 14496-4 Amd.24

28.3.1 Topics1. File Format Conformance

68

28.3.2 ContributionsM14487: Contribution to Conformance for ISO/IEC 14496-12 AMD.1. Accepted and integrated in the text of the FPDAM.M14289: Summary of Voting on ISO/IEC 14496-4:2004/PDAM 24 [SC 29 N 8182]. All comments have been disposed of. See DoC.

-- only FR boiler-plate comment-- see 8648 (Hangzhou)

Updated with one new file (timed meta-data), from Michael.

Technical Work in Progress.

28.4 14496-4 Amd.25 LASeR V1 Conformance

28.4.1 Topics1. LASeR Conformance

28.4.2 ContributionsM14290: Summary of Voting on ISO/IEC 14496-4:2004/PDAM 25 [SC 29 N 8184]. All comments have been disposed of. See DoC.

Technical Work in Progress.

28.5 14496-4 Amd.26 Open Font Format Conformance

28.5.1 Topics2. Open Font Format Conformance

28.5.2 ContributionsM14402: Proposed conformance test methodology and bitstreams for ISO/IEC 14496-22. Taken as the basis for the production of the PDAM.

Technical Work in Progress.

28.6 14496-4 Amd.27 LASeR V2 Conformance

28.6.1 Topics1. LASeR V2 Conformance

28.6.2 ContributionsNone.

Technical Work in Progress.

29 MPEG-4 Reference Software (14496-5)

29.1 14496-5 Amd.12

29.1.1 Topics1. ISO File Format Reference Software

29.1.2 ContributionsM14324 : Summary of Voting on ISO/IEC 14496-5:2001/FPDAM 12 [SC 29 N 8273]. All comments have been disposed of. See DoC.

69

-- only FR boiler-plate comment-- see 8653 (Hangzhou)

Updated with bug fixes, more support for 3G and better sample entry support, from Dave.

Technical Work Finalized.

29.2 14496-5 Amd.14

29.2.1 Topics1. Open Font Format Reference Software

29.2.2 ContributionsNone

Technical Work in Progress.

29.3 14496-5 Amd.16

29.3.1 Topics1. Symbolic Music Representation Reference Software

29.3.2 ContributionsNone.

Technical Work in Progress.

29.4 14496-5 Amd.17

29.4.1 Topics1. LASeR Reference Software

29.4.2 ContributionsNone.

Technical Work in Progress.

30 Scene Representation (14496-11)

30.1 14496-11:2005 Cor.6

30.1.1 Topics1. AudioFX Proto

30.1.2 ContributionsM14387: Summary of Voting on ISO/IEC 14496-11:2005/DCOR 6. No comment. COR produced.

Technical Work Finalized.

70

31 ISO File Format (14496-12)

31.1 14496-12/Amd.2

31.1.1 Topics1. Flute Hint Track

31.1.2 14122 ISO Base Media File Format Branding14336 Summary of Voting on ISO/IEC 14496-12:2005 2 & ISO/IEC 15444-12:2005/FPDAM 2. US and SE comments only. See the disposition of comments report. We have a potential issue wrt referring to I-Ds.

14404 Comments and suggestions regarding ISO/IEC 14496-12 Amd.2. Thank you for the careful read and the editorial improvements.

31.2 Miscellaneous

14529: MP4 file format considerations for high sample-rate audio. We will see what the conformance files do, but perhaps the Corr. we issued in Marrakech is enough.

14525: Signaling of leading pictures in file format. This is neat. We like it. But we think we can fit it into the sample dependency table, the reserved two bits (which are also available in the movie fragments). At this meeting we propose a new output document “technologies under consideration” for Part 12, in which we hope to collect other amendment-ready material and then issue it sometime soon. “is, is-not, unknown, reserved” leading picture, where leading picture is defined with respect to the previous sample marked as an I picture,

Technical Work Finalized.

31.3 14496-12/Cor.3

31.3.1 Topics1. Misc. Correction on File Format

31.3.2 14264 AAC SBR timescales and sample ratesM14388: Summary of Voting on ISO/IEC 14496-12:2005/DCOR 3 & ISO/IEC 15444-12:2005/DCOR 3. No comment. COR produced.

-- 19 approve, no disapprove.-- dealt with audio fields in MP4 files (8873)

Technical Work Finalized.

71

32 MPEG-4 AVC File Format (14496-15)

32.1 14496-15:2004/Amd.2

32.1.1 Topics1. SVC File Format Extensions

14405: Comments on the SVC File Format. Thank you. Text adjusted.

14494: Extraction path description. This seems interesting, but also quite complex. It is interesting to try to describe extraction paths and their consequences, but we’re not sure of the description.

14495: Terms and definitions for the SVC file format . Excellent, thank you.

14496: On the SVC File format. Yes, extractors need to be temporally mis-aligned, and we agree to have a sample offset (+/- sample count), and be careful about defining temporally aligned. Yes, we need to adjust for prefix/suffix, and for FGS, tl0 etc. Thank you, the toolsets adjusted also. On ROI, we understand the desire, but it does seem a little integration and description work may be needed. E.g., how do I know what ‘object’ each ROI is tracking?

14526: On SVC file format. HRD done, thank you! The quality information goes next to the scalabilityinfoSEIbox (or maybe a Tier); we choose the first for now. We don’t link it to meta-data or anything.We don’t think we need a slice header meta-data statement (yet). We would welcome re-timing information possibly using sample groups, and/or time-parallel meta-data, or new boxes in the sample table.

14527: Signaling of temporal layer switching points in SVC file format. Accepted, thank you.

14550: Addendum to ISO/IEC 14496-15 AMD2: File Format Support for Scalable Video Coding. Yes, version 1 of sample groups is required. JVT Joint Meeting. The SVC specification will formally be a study text, produced on a long editing period, from this meeting. That study should be available 2-3 weeks before the next meeting. It is intended that a minor variant of that will be approved as the final text, at the next meeting. Given that the high-level syntax is still not firm, it would be imprudent for the file format text to go to ballot at this meeting. In particular, the exact NAL types, and the use of prefixes, are still under discussion.

32.2 JPSEC/FFSEC Joint meetingWe had a good exchange of designs and the motivation for those designs. We’re going to encourage FFSEC people to join the MP4 reflector, and in email correspondence between now and Lausanne work on the aspects of the FFSEC design that could be more general. At some point these pieces could (should) be moved into Part 12 and Corr’d out of FFSEC, but they can start in FFSEC.

We also shared some information on IPMP.

We think it is too late to make significant changes to the FLUTE amendment.

Areas that look fruitful include:

72

a) general design for layered protection (asked for at this meeting by an MPEG-21 person)

b) some kind of item reference box, like a track reference box (typed references), that would allow for ‘annotation’ or linking of items

c) a better design than the ‘xml box’ for putting item data inside the meta-boxd) maybe some kind of sub-item information/structure boxe) some kind of support for general scalability, not codec specific (SVC extractors are specific to

SVC); perhaps also a ‘scalable RTP hint track’

There may be other areas.

32.3 MDS Joint meetingWe considered the input contribution M14365, and a number of ideas were raised. First, it is possible to use the item protection provisions at file format level as well as at the DIDL level. Second, it is possible to embed a digital item as a file item resource to another digital item, and protect it whole. Third, the layered protection design being done with FFSEC (above) might help in future. There did seem to be a tension between including something ‘as an item’, and wanting to protect it ‘as an item’, yet still wanting to see its structure.

Technical Work in Progress.

33 LASeR (14496-20)

33.1 14496-20/Amd. 1

33.1.1 Topics1. Lightweight Application Scene Representation (LASeR Extensions)

33.1.2 ContributionsM14373: LASeR profiles adjustments. Accepted after discussion and integrated into the FDAM.M14551: Proposal for a new LASeR Profile. after discussion and integrated into profile under consideration.

Technical Work Finalized.

33.2 14496-20/Amd. 2

33.2.1 Topics1. Lightweight Application Scene Representation (SGVT1.2 Support)

33.2.2 ContributionsM14372 This contribution proposes split of current AMD in two pieces since SVGT1.2, the technology LASeR scene description is based on, does not seem to be finished by July meeting. Therefore, the elements not related to SVGT1.2 are promoted to FPDAM at this meeting. And new AMD will start this meeting to hold remainders.

M14370 This contribution proposes changes to AMD1 for harmonization between LASeR and 3GPP DIMS.

73

Additional width and height fields for rectClip containing same values with what the size field represents. (if two values doesn’t match, the last values will be used.)

Renaming of updateSource to updates and addition of syncReference Reduce the cases of rotation by two because orientation of the screen is always the top

left corner of the resulted screen by rotation. (portrait or landscape is only matters) no semantic changes but names.

Proposed modification will be implemented in AMD1

M14378 This contribution lists new technologies coming from 3GPP Immediate Script Execution for script executed immediately without inserting script

node and removed after execution New command, “seek”, for seeking across the boundary of presentation regardless of

the scene time to be reset at the execution of NewScene command.Proposed technologies will be included in the AMD2.

M14418 This contribution analysis the relationship between MPEG-21 and LASeR. It is identified converting DID into LASeR for presentation cannot be done easily. So it is proposed to use LASeR as a presentation description for DI and this is agreed with MDS subgroup during the joint meeting. It is decided to include this contribution in the output document about Items under considerations in LASeR.

M14419 This contribution analysis the possibilities and the potential issues of carrying ISO/IEC 14496-20 contents over MPEG-2. It is decided to include this contribution in the output document about Items under considerations in LASeR.

Technical Work in Progress.

33.3 14496-20/Cor 2

33.3.1 Topics1. Profile Removal

33.3.2 ContributionsNone.

Technical Work in Progress.

34 21000-09 MPEG-21 File Format

34.1 MPEG-21 File Format Amendment

34.1.1 Topics1. Mime Type

34.1.2 ContributionsM14555: MIME Type registration for MPEG-21 File Format. Accepted and used as the basis for the production of the PDAM text.

74

35 21000-14 Conformance

35.1 MPEG-21 File Format Conformance

35.1.1 Topics1. Conformance

35.1.2 ContributionsM14497: French NB comment on FCD 21000-14. All comments have been addressed. See DoC.M11451: Binary Conformance streams for MPEG-21. Accepted. Integrated in text of FDIS.

36 MPEG-A MAF (23000)

36.1 23000-4 Musical Slide Show MAF

36.1.1 Topics2. Musical Slide Show MAF

36.1.2 ContributionsM14343: Summary of Voting on ISO/IEC FCD 23000-4 [SC 29 N 8306]. All comments have been disposed of. See DoC.

M14437: A proposal on metadata modification for Musical Slide Show MAF. Accepted. Will be included in FDIS.

Technical Work Finalized

36.2 23000-8 Portable Video Player MAF

36.2.1 Topics1. Portable Video Player MAF

36.2.2 Contributions14435: Proposed text of ISO/IEC 23000-8 CD Portable video player MAF. Taken as a basis for the documentation of the CD.

14438 : A proposal of an additional functionality to be supported in Portable Video Player MAF. Accepted.

Technical Work in Progress.

36.3 23000-9 Digital Multimedia Broadcasting MAF

36.3.1 Topics1. Digital Multimedia Broadcasting MAF

36.3.2 ContributionsM14394 : Summary of Voting on ISO/IEC CD 23000-9. All comments have been addressed and documented in DoC.

M14425: (Editors Input) Updated Text of ISO/IEC 23000-9 MAF for DMB. Taken as input for producing text of FCD.

75

M14426 This contribution proposes a method to store MPEG-2 TS in a MP4 file. It was identified during the discussion that DVB is working on the same problem. So it is decided to send a liaison letter to DVB and to try to find the harmonized solution before we take a specific solution for this MAF. Proposed method will be included in the TuC.

M14427 This contribution present the draft list of TVA features appropriate to be used in MAF for DMB. Since the selection is not completed and the schema is not validated yet, this will be included in the Technologies under consideration.

Technical Work in Progress.

37 MPEG-B

37.1 23001-1 Binary Format Amd.2

37.1.1 Topics1. Extension on Encoding of Wild Cards

37.1.2 ContributionsM14450: Editor's study of 23001-1 FPDAM2. Taking into account for the production of study text.

Technical Work in Progress.

37.2 23001-1 Binary Format Cor.2

37.2.1 Topics1. Misc. Editorial Corrections on MPEG-B Part 1

37.2.2 ContributionsM14395 : Summary of Voting on ISO/IEC 23001-1:2006/DCOR 2. See DoC.

Technical Work in Progress.

37.3 23001-2 Fragment Request Unit

37.3.1 Topics1. Fragment Request Unit

37.3.2 ContributionsM14381: Summary of Voting on ISO/IEC FCD 23001-2. See Doc.

Technical Work Finalized

37.4 23001-3 Binary to XML Mapping of IPMP-X

37.4.1 Topics1. Binary to XML Mapping of IPMP-X

37.4.2 ContributionsM14299: Summary of Voting on ISO/IEC CD 23001-3 [SC 29 N 8227]. No comments. Text of FCd was produced.

76

M14443: Proposed text of ISO/IEC 23001-3 FCD Binary XML to IPMP-X. Taken as input for the production of the FCD.

M14498: Proposal of Modified IPMP XML messages for ISO/IEC 23001-3 Binary XML to IPMP-X. Approved and included in the FCD.

Technical Work in Progress.

38 MPEG-E Multimedia Middleware (23004)

38.1 Multimedia Middleware

38.1.1 Topics1. MPEG Multimedia Middleware

38.1.2 ContributionsAt the 80th MPEG Meeting in San Jose, California, USA (April 23 – 27, 2007) MPEG has promoted the remaining three parts (Part 5: Component Download, Part 6: Fault Management and Part 7: System Integrity Management) of M3W (ISO/IEC 23004, MPEG-E (Multimedia Middleware)) to the FDIS (Final Draft International Standard) stage. Please note that the first four parts (Part 1: Architecture, Part 2: Multimedia API, Part 3: Component Model and Part 4: Resource and Quality Management) have already reached this stage at the previous MPEG meeting in January 2007. This implies that all seven parts of M3W are now completed.

At the April MPEG Meeting also a second version of the WD (Working Draft) for the reference software and conformance testing (Part 8: Reference Software and Conformance) has been released. The reference software and conformance testing includes the  implementation of the logical components and optional frameworks, supporting tools and sample application demonstrating the functionality of the individual parts and this then feeds in to the conformance testing process. The associated plan for the delivery of the reference software and conformance testing (M3W Reference Software and Conformance Plan) has been updated to reflect the current status and future planned activities.”

M14337 : Summary of Voting on ISO/IEC FCD 23004-5 [SC 29 N 8298]. See DoC.M14338 : Summary of Voting on ISO/IEC FCD 23004-6 [SC 29 N 8299]. See DoC.M14339 : Summary of Voting on ISO/IEC FCD 23004-6 [SC 29 N 8299]. See DoC.

M14371 : Contribution to M3W Reference Software for M3W Parts 2, 3, 5, 6 & 7. Taken as input for the production of WD2.0 of ISO/IEC 23004-8 Reference Software and Conformance.

Technical Work in Progress.

77

39 Supplementary Media Technology (29116-1)

39.1 Media Streaming MAF Protocols

39.1.1 Topics1. Media Streaming MAF Protocols

39.1.2 ContributionsM14304: Summary of Voting on ISO/IEC CD 23005-1 [SC 29 N 8236]. All comments have been disposed of. See DoC.

M14460: Austrian NB comments on ISO/IEC CD XXXXX Media Streaming MAF Protocols. See DoC.

M14444: Proposed text of ISO/IEC 23005-1 FCD Media Streaming MAF Protocol (Editor's Input). Taken as input to produce text of the FCD.

Technical Work in Progress.

40 Exploration

M14418: Ideas on MPEG-21 and LASeR. Follow-up on the discussion we had in previous meeting.1. Exploration on convertion of digital items to LASeR MPEG-21 into LASeR. Hard

to convert MPEG-21 into LASeR. 2. Add LASeR representation in MPEG-21 and a DIBO.

Document on first ideas on LASeR was updated.

78

41 Latest References and Publication Status

Pr Pt Standard No. Issue Status Doc. With Purpose ISO Award

2 1 ISO/IEC 13818-1/Amd.7 Published 2000/12 Done2 1 ISO/IEC 13818-1:2000 (MPEG-2 Systems 2nd Edition) 00/12 Published 2000/12 Proposed2 1 ISO/IEC 13818-1:2000/COR1 (FlexMux Descr.) N3844 01/01 Pisa Published 2002/03 N/A2 1 ISO/IEC 13818-1:2000/COR2 (FlexMuxTiming_ descriptor) N4404 01/12 Pattaya Published 2002/12 N/A2 1 ISO/IEC 13818-1:2000/Amd.1 (Metadata on 2) & COR1 on Amd.1 N5867 03/07

TrondheimPublished 2003/12 Proposed

2 1 ISO/IEC 13818-1:2000/Amd.2 (Support for IPMP on 2) N5604 03/03 Pattaya Published 2004/03 N/A2 1 ISO/IEC 13818-1:2000/Amd.3 (AVC Carriage on MPEG-2) N5771 03/07

TrondheimPublished XXXX Proposed

2 1 ISO/IEC 13818-1:2000/Amd.4 (Metadata Application CP) N6847 04/10 Palma FDAM ITTF to be published N/A2 1 ISO/IEC 13818-1:2000/Amd.5 (New Audio P&L Sig.) N6585 04/07

RedmondFDAM ITTF to be published N/A

2 1 ISO/IEC 13818-1:2000/COR3 (Correction for Field Picture) N6845 04/10 Palma COR ITTF to be published N/A2 1 ISO/IEC 13818-1:2000/COR4 (M4MUX Code Point) N7469 05/07 Poznan COR ITTF to be published N/A2 1 ISO/IEC 13818-1:2000/COR5 (Corrections related to 3rd Ed.) N7895 06/01

BangkokCOR ITTF to be published N/A

2 1 ISO/IEC 13818-1:2006 (MPEG-2 Systems 3rd Edition) 06/xx Published ITTF TBP2 1 ISO/IEC 13818-1:2006/Amd.1 (Transport of Streaming text) N8369 06/07

KlagenfurtFDAM ITTF to be published TBP

2 1 ISO/IEC 13818-1:2006/Amd.2 (Carriage of Auxialiry Video Data) N8798 07/01 Marrakech

FDAM ITTF to be published TBP

2 11 ISO/IEC 13818-1:2003 (IPMP on 2) N5607 03/03 Pattaya Published 2003/12 Proposed4 1 ISO/IEC 14496-1 (MPEG-4 Systems 1st Ed.) N2501 98/10 Atl. City Published 1999/12 Done4 1 ISO/IEC 14496-1/Amd.1 (MP4, MPEG-J) N3054 99/12 Hawaii Published 2001/11 Done4 1 ISO/IEC 14496-1/Cor.1 N3278 00/03 Published 2001/11 N/A

79

Noordwijk.4 1 ISO/IEC 14496-1:2001 (MPEG-4 Systems 2nd Ed.) N3850 01/01 Pisa Published 2001/11 N/A4 1 ISO/IEC 14496-1:2001/Amd.1 (Flextime) Published 2002/10 Done4 1 ISO/IEC 14496-1:2001/Cor.1 N4264 01/07 Sydney COR ITTF N/A4 1 ISO/IEC 14496-1:2001/Cor.2 N5275 02/10 Shangai COR ITTF N/A4 1 ISO/IEC 14496-1:2001/Cor.3 N6587 04/07

RedmondCOR ITTF N/A

4 1 ISO/IEC 14496-1:2001/Amd.2 (Textual Format) N4698 02/03 Jeju Island

AMD ITTF N/A

4 1 ISO/IEC 14496-1:2001/Amd.3 (IPMP Extensions) N5282 02/10 Shanghai

Published 2004-05 N/A

4 1 ISO/IEC 14496-1:2001/Amd.4 (SL Extension) N5471 02/12 Awaji Published 2003/12 N/A4 1 ISO/IEC 14496-1:2001/Amd.7 (AVC on 4) N5976 03/10

BrisbannePublished 2004-08 N/A

4 1 ISO/IEC 14496-1:2001/Amd.8 (ObjectType Code Points) N6202 03/12 Hawaii AMD ITTF to be published N/A4 1 ISO/IEC 14496-1:200x/Amd.1 (Text Profile Descriptors) N7229 05/04 Busan PDAM ITTF Final Text

EditingN/A

4 1 ISO/IEC 14496-1:200x/Cor4 (Node Coding Table) N7473 05/07 Poznan PDAM ITTF to be published N/A4 1 ISO/IEC 14496-1 (MPEG-4 Systems 3rd Ed.) N5277 02/10

ShanghaiIS ITTF to be published Proposed

4 1 ISO/IEC 14496-1:200x/Amd.1 (Text Profile Descriptors) N7229 05/04 Busan PDAM ITTF Final Text Editing

N/A

4 1 ISO/IEC 14496-1:200x/Cor.1 (Clarif. On audio codec behavior) N8117 06/04 Montreux

COR ITTF Final Text Editing

N/A

4 1 ISO/IEC 14496-1:200x/Amd.2 (3D Profile Descriptor Extensions) N8372 06/07 Klagenfurt

PDAM ITTF to be published N/A

4 1 ISO/IEC 14496-1:200x/Cor.2 (OD Dependencies) N8646 06/10 Hangzhou

COR ITTF to be published N/A

4 1 ISO/IEC 14496-1:200x/Amd.3 (JPEG 2000 support in Systems) N8860 07/01 Marrakech

PDAM ITTF to be published N/A

4 4 ISO/IEC 14496-1:200x/Amd.17 (ATG Conformance) N8861 07/01 Marrakech

PDAM ITTF to be published N/A

80

4 5 ISO/IEC 14496-1:200x/Amd.12 (File Format) N9020 07/04 San Jose PDAM ITTF to be published N/A4 6 ISO/IEC 14496-6:2000 Published 2000/12 N/A4 8 ISO/IEC 14496-8 (MPEG-4 on IP Framework) N4712 02/03 Jeju Published 2004-05 Proposed4 11 ISO/IEC 14496-11 (MPEG-4 Scene Description 3rd

Edition) N6960 05/01

HongKongFDIS SC29 Final Text

EditingProposed

4 11 ISO/IEC 14496-11/Amd.1 (AFX) N5480 02/12 Awaji FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Amd.2 (Advanced Text and Graphics) N6205 03/12 Hawaii FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Cor.1 N6203 03/12 Hawaii COR SC29 N/A4 11 ISO/IEC 14496-11/Cor.3 Valuator/AFX related correction N6594 04/07

RedmondCOR ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Amd.3 Audio BIFS Extensions N6591 04/07 Redmond

FDAM ITTF Integration in 1st

Ed.Proposed

4 11 ISO/IEC 14496-11/Amd.4 XMT and MPEG-J Extensions N6959 05/01 HongKong

FDAM ITTF Integration in 1st

Ed.N/A

4 11 ISO/IEC 14496-11/Cor.3 (Audio BIFS Integrated in 3rd Edition) N7230 05/04 Busan COR ITTF Final Text Editing

N/A

4 11 ISO/IEC 14496-11/Cor.5 (Misc Corrigendum) N8383 06/07 Klagenfurt

COR SC29 N/A

4 11 ISO/IEC 14496-11/Amd.5 Symbolic Music Representation

N8657 06/10 Hangzhou

FDAM ITTF TBP

4 11 ISO/IEC 14496-11/Cor.6 (AudioFx Correction) N9021 07/04 San Jose COR SC29 N/A4 12 ISO/IEC 14496-12 (ISO Base Media File Format) N5295 02/10

ShanghaiPublished 2004-02 Proposed

4 12 ISO/IEC 14496-12/Amd.1 ISO FF Extension N6596 04/07 Redmond

FDAM ITTF FDAM 04/11/30 N/A

4 12 ISO/IEC 14496-12/Cor.1 (Correction on File Type Box)

N7232 05/04 Busan COR ITTF Final Text Editing

N/A

4 12 ISO/IEC 14496-12/Cor.2 (Miscellanea) N7901 06/01 Bangkok

COR ITTF Final Text Editing

N/A

4 12 ISO/IEC 14496-12/Cor.3 (Miscellanea) N9024 07/04 San Jose COR ITTF Final Text N/A

81

Editing4 12 ISO/IEC 14496-12/Amd.1 (Description of timed

metadata) N8659 06/10

HangzhouFDAM ITTF N/A

4 12 ISO/IEC 14496-12/Amd.2 (Flute Hint Track) N9023 07/04 San Jose FDAM ITTF N/A4 13 ISO/IEC 14496-13 (IPMP-X) N5284 02/10

ShanghaiIS ITTF to be published Proposed

4 14 ISO/IEC 14496-14 (MP4 File Format) N5298 02/10 Shanghai

Published 2003-11 Proposed

4 14 ISO/IEC 14496-14/Cor.1 (Audio P&L Indication) N7903 06/01 Bangkok

COR ITTF Final Text Editing

N/A

4 15 ISO/IEC 14496-15 (AVC File Format) N5780 03/07 Trondheim

Published 2004-04 Proposed

4 15 ISO/IEC 14496-15/Amd.1 (Support for FREXT) N7585 05/10 Nice FDAM ITTF Final Text Editing

N/A

4 15 ISO/IEC 14496-15/Cor.1 N7575 05/10 Nice COR ITTF N/A4 15 ISO/IEC 14496-15/Cor.2 (NAL Unit Restriction) N8387 06/07

KlagenfurtCOR ITTF N/A

4 17 ISO/IEC 14496-17 (Streaming Text) N7479 05/07 Poznan FDAM ITTF TBP4 18 ISO/IEC 14496-18 (Font Compression and Streaming) N6215 03/12 Hawaii Published 2004-07 Proposed4 18 ISO/IEC 14496-18/Cor.1 (Misc. corrigenda and

clarification)N8664 06/10

HangzhouCOR ITTF N/A

4 19 ISO/IEC 14496-19 (Synthesized Texture Stream) N6217 03/12 Hawaii Published 2004-07 Proposed4 20 ISO/IEC 14496-20 (LASeR) N7588 05/10 Nice FDAM Editor TBP4 20 ISO/IEC 14496-20/Cor.1 (Misc. corrigenda and

clarification)N8666 06/10

HangzhouCOR ITTF N/A

4 20 ISO/IEC 14496-20/Amd.1 (LASeR Extension) N9029 07/04 San Jose FDAM ITTF N/A4 22 ISO/IEC 14496-22 (Open Font Format) N8395 06/07

KlagenfurtFDAM Editor Final Text

EditingTBP

7 1 ISO/IEC 15938-1 (MPEG-7 Systems) N4285 01/07 Sydney Published 2002/07 Done7 1 ISO/IEC 15938-1/Amd.1 (MPEG-7 Systems Extensions) N6326 04/03 Munich FDAM ITTF FDAM 04/11/28 N/A7 1 ISO/IEC 15938-1/Cor.1 (MPEG-7 Systems Corrigendum) N6328 04/03 Munich COR Editor N/A7 1 ISO/IEC 15938-1/Cor.2 (MPEG-7 Systems Corrigendum) N7490 05/07 Poznan COR ITTF N/A

82

7 1 ISO/IEC 15938-1/Amd.2 (BiM extension) N7532 05/10 Nice FDAM ITTF N/A7 2 ISO/IEC 15938-2 (MPEG-7 DDL) N4288 01/07 Sydney Published 2002/02 Done7 7 ISO/IEC 15938-7/Amd.2 (Fast Access Ext. Conformance) N8672 06/10

HangzhouFDAM ITTF N/A

21 9 ISO/IEC 21000-9 (MPEG-21 File Format) N6975 05/01 HongKong

FDIS ITTF FDIS 05/01/21 Done

21 16 ISO/IEC 21000-16 (MPEG-21 Binary Format) N7247 05/04 Busan FDIS ITTF FDIS 05/04/22 TBPA 1 ISO/IEC 23000-4 (Musical Slide Show MAF) N9037 07/04 San Jose FDIS ITTF TBPB 1 ISO/IEC 23001-1 (XML Binary Format) N7597 05/10 Nice FDIS ITTF TBPB 1 ISO/IEC 23001-1/Cor.1 (Misc. Editorial and technical

clar.)N8680 06/10

HangzhouCOR ITTF N/A

B 1 ISO/IEC 23001-1/Cor.2 (Misc. Editorial and technical clar.)

N9049 07/04 San Jose COR ITTF N/A

B 1 ISO/IEC 23001-1/Amd.1 (Reference Soft. & Conf.) N8886 07/01 Marrakech

FDAM ITTF N/A

B 2 ISO/IEC 23001-1 (Fragment Request Unit) N9051 07/04 San Jose FDIS ITTF TBPE 1 ISO/IEC 23008-1 Architecture N8892 07/01

MarrakechFDAM ITTF N/A

E 2 ISO/IEC 23008-2 Multimedia API N8893 07/01 Marrakech

FDAM ITTF N/A

E 3 ISO/IEC 23008-3 Component Model N8894 07/01 Marrakech

FDAM ITTF N/A

E 4 ISO/IEC 23008-4 Ressource & Quality Management N8895 07/01 Marrakech

FDAM ITTF N/A

E 5 ISO/IEC 23008-5 Component Download N9053 07/04 San Jose FDAM ITTF N/AE 6 ISO/IEC 23008-6 Fault Management N9054 07/04 San Jose FDAM ITTF N/AE 7 ISO/IEC 23008-7 System Integrity Management N9055 07/04 San Jose FDAM ITTF N/A

83

42 Resolutions of Systems

Cf. WG11 resolution.

43 List of Reviewed Contributions

N° Title Authors14289 Summary of Voting on ISO/IEC

14496-4:2004/PDAM 24 [SC 29 N 8182]SC 29 Secretariat

14290 Summary of Voting on ISO/IEC 14496-4:2004/PDAM 25 [SC 29 N 8184]

SC 29 Secretariat

14297 Liaison Statement from 3GPP [SC 29 N 8225] 3GPP via SC 29 Secretariat

14299 Summary of Voting on ISO/IEC CD 23001-3 [SC 29 N 8227]

SC 29 Secretariat

14304 Summary of Voting on ISO/IEC CD 23005-1 [SC 29 N 8236]

SC 29 Secretariat

14305 Liaison Statement from the DVD Forum WG-1 [SC 29 N 8254]

the DVD Forum WG-1 via SC 29 Secretariat

14324 Summary of Voting on ISO/IEC 14496-5:2001/FPDAM 12 [SC 29 N 8273]

SC 29 Secretariat

14329 USNB Contribution: Response to resolution 3.1.2 of 79-th WG 11 meeting

A. G. Tescher for USNB

14336 Summary of Voting on ISO/IEC 14496-12:2005/FPDAM 2 and ISO/IEC 15444-12:2005/FPDAM 2 [SC 29 N 8297]

SC 29 Secretariat

14337 Summary of Voting on ISO/IEC FCD 23004-5 [SC 29 N 8298]

SC 29 Secretariat

14338 Summary of Voting on ISO/IEC FCD 23004-6 [SC 29 N 8299]

SC 29 Secretariat

14339 Summary of Voting on ISO/IEC FCD 23004-7 [SC 29 N 8301]

SC 29 Secretariat

14343 Summary of Voting on ISO/IEC FCD 23000-4 [SC 29 N 8306]

SC 29 Secretariat

14349 Liaison re w8559 Text of ISO/IEC 13818-1:200x/DCOR.1

Gavin SchutzTeruhiko SuzukiMichael Dolan

14362 Liaison Statement from the DVB [SC 29 N 8326] DVB via SC 29 Secretariat

14366 Additional examples on Cross-Media Interactive Presentation MAF

Paolo NesiPierfrancesco Bellini Davide Rogai

14367 Proposal for a MAF on Cross-Media Interactive Presentation: Application Scenarios

Paolo NesiPierfrancesco Bellini Davide RogaiKia Ng (University of

84

N° Title AuthorsLeeds)

14368 Proposal for a MAF on Cross-Media Interactive Presentation: Requirements

Paolo NesiPierfrancesco Bellini Davide Rogai

14369 Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFs

Davide RogaiPierfrancesco BelliniPaolo Nesi

14370 LASeR fixes requested by 3GPP DIMS Jean-Claude Dufourd14371 Contribution to M3W Reference Software for

M3W Parts 2, 3, 5, 6 & 7Jean H.A. Gelissen (editor)Johan Muskens

14372 Splitting LASeR AMD1 Jean-Claude Dufourd14373 LASeR profiles adjustments Jean-Claude Dufourd14378 Additions to LASeR AMD2 from 3GPP Jean-Claude Dufourd14381 Summary of Voting on ISO/IEC FCD 23001-2 SC 29 Secretariat14382 Summary of Voting on ISO/IEC

13818-1:200X/DCOR 1SC 29 Secretariat

14385 Summary of Voting on ISO/IEC 14496-4:2004/PDAM 23

SC 29 Secretariat

14387 Summary of Voting on ISO/IEC 14496-11:2005/DCOR 6

SC 29 Secretariat

14388 Summary of Voting on ISO/IEC 14496-12:2005/DCOR 3 & ISO/IEC 15444-12:2005/DCOR 3

SC 29 Secretariat

14394 Summary of Voting on ISO/IEC CD 23000-9 SC 29 Secretariat14395 Summary of Voting on ISO/IEC

23001-1:2006/DCOR 2SC 29 Secretariat

14402 Proposed conformance test methodology and bitstreams for ISO/IEC 14496-22

Simon DanielsVladimir Levantovsky

14404 Comments and suggestions regarding ISO/IEC 14496-12 Amd.2

Jani PeltotaloMiska M. Hannuksela

14405 Comments on the SVC File Format David Singer14413 Liaison Statement from TTA [SC 29 N 8333] TTA via SC 29

Secretariat14418 Ideas on MPEG-21 and LASeR Jihun Cha

YeSun JoungYoung-Kwon LimKyungAe Moon

14418 Ideas on MPEG-21 and LASeR Jihun ChaYeSun JoungYoung-Kwon LimKyungAe Moon

14419 Issues on the carriage of ISO/IEC 14496-20 contents over MPEG-2

Jihun ChaYoungkwon LimYeSun JoungKyungAe Moon

14425 (Editors Input) Updated Text of ISO/IEC 23000-9 MAF for DMB

Hui Yong KimHyon-Gon Choo Munchurl Kim

85

N° Title Authors14426 Proposal for MPEG-2 TS Encapsulation with

ISO/IEC 23000-9 MAF for DMBHui Yong KimGun BangMyungSeok KiHyun Cheol KimHan-Kyu LeeJin Woo HongYoung-Kwon Lim

14427 Proposal for Restrictions on TV-Anytime Metadata in ISO/IEC 23000-9 MAF for DMB

Hui Yong KimSeung Jun YangHeekyung LeeHan-Kyu LeeJin Woo HongMunchurl KimJinhan Kim

14435 Proposed text of ISO/IEC 23000-8 CD Portable video player MAF

Hyouk Jean ChaTae Hyeon KimHerbert Thoma

14437 A proposal on metadata modification for Musical Slide Show MAF

Ryoma Oami

14438 A proposal of an additional functionality to be supported in Portable Video Player MAF

Ryoma Oami

14443 Proposed text of ISO/IEC 23001-3 FCD Binary XML to IPMP-X

Filippo Chiariglione(Editor)Hyon-Gon Choo(Editor)Jooyoung Lee

14444 Proposed text of ISO/IEC 23005-1 FCD Media Streaming MAF Protocol (Editor's Input)

Hyon-Gon ChooFilippo ChiariglioneNaito Joji

14450 Editor's study of 23001-1 FPDAM2 David Thevenin Philippe de Cuetos

14451 Binary Conformance streams for MPEG-21 David Thevenin Philippe de Cuetos

14460 Austrian NB comments on ISO/IEC CD XXXXX Media Streaming MAF Protocols

14477 Updated Proposal for Protected Musical Slide Show MAF with IPMP

HendryHouari SabirinMunchurl Kim

14478 Updated Proposal for Protected Photo Player MAF with IPMP

HendryHouari SabirinMunchurl Kim

14487 Contribution to Conformance for ISO/IEC 14496-12 AMD/1

Michael RansburgHermann Hellwagner

14489 Proposal of Modified IPMP XML messages for ISO/IEC 23001-3 Binary XML to IPMP-X

Filippo ChiariglioneJooyoung LeeHyon-Gon Choo

14494 Extraction path description Thomas RathgenMichael RansburgPeter AmonAndreas Hutter

86

N° Title AuthorsHermann Hellwagner

14495 Terms and definitions for the SVC file format Michael RansburgThomas RathgenPeter AmonAndreas HutterHermann Hellwagner

14496 On the SVC file format Thomas RathgenPeter AmonAndreas Hutter

14497 French NB comment on FCD 21000-14 Philippe de Cuetos on behalf of FNB

14512 Proposed technical alternative to MPEG-2 Systems DCOR 1 text WG 11 N 8859

Gary J. SullivanRegis Crinon

14525 Signaling of leading pictures in file format Ying Chen Ye-Kui Wang Miska M. Hannuksela

14526 On SVC file format Ye-Kui Wang Miska M. Hannuksela

14527 Signaling of temporal layer switching points in SVC file format

Ye-Kui Wang Miska M. Hannuksela

14529 MP4 file format considerations for high sample-rate audio

David Singer

14535 Liaison Statement from JSR 287 Expert Group [SC 29 N 8336]

JSR-287 EG via SC 29 Secretariat

14551 Proposal for a new LASeR Profile Jean-Claude Dufourd

14555 MIME Type registration for MPEG-21 File Format

87

Annex G – MDS report

Source: Ian Burnett, Chair

1.0 Introduction

MDS commenced with an overview of the weeks planned activities:

88

89

90

2.0 Notes on discussions on Input DocumentsThese contemporaneous notes summarise the activities of the MDS subgroup during the 80 th MPEG meeting. Over the week several short break out activities dealt with tasks. The Break out groups worked on the REL and Open Release MAF, Professional Archival work and MPEG-7 Query. (Reports of the break out groups are included at the end of this section).

Following a short MPEG plenary, a joint meeting with Requirements considered the following from 11am -12pm.14475 Giovanni Cordara (on behalf of the ITNB) Italian NB proposal to revisit MPEG-21 DID

Input:This input from the Italian NB proposes the development of a new DID based on a royalty-free basis.

91

Actions:There appears to be support for a royalty free standard in MPEG but one question is whether MPEG can usefully create a new DID standard.Issues:

1. Royalty Free2. Technical Issues – agnosticism of the DI, application specific containers

A BOG was established to specifically consider the technical issues. The discussion will consider the limitations of the current DID as a starting pointMDS officially opened at 1.30pm with a run through of the weeks activities. Note that in the following only MDS input documents are discussed. Joint meetings with Requirements and Systems (see schedule) and the treatment of those documents considered in the joint meetings are considered in the respective group reports.14415 Kisong Yoon Taehyun Kim Hogab Kang

Interoperability between MPEG-21 REL DAC Profile and Other Standards

Input:This input considers how the DAC profile will provide interoperability with TV-Anytime, DVB and OMA.

Actions:There was agreement that this was a very useful analysis. MDS will investigate ways to publicise this information.14484 Kisong Yoon Taehyun Kim Hogab Kang

A Study on Use Cases of Derivative Works with MPEG-21 REL ORC Profile License

Input:This input considers how the ORC profile will provide for derivative and aggregate works.

Actions:The input proposes structures for licenses to provide effectively for derivative works. Thiswas discussed further in the BOG14507 Eva Rodríguez Jaime Delgado

Contribution to the current version of the Open Release MAF

Input:This input considers how the OR MAF and suggests addition of descriptions of DIDL elements etc. It wants the text to make the usage of the elements more specific.

Actions:Discussions suggested that profiling of the DID wasn’t a solution. There were questions raised as to why full DID descriptions were needed. It seems that it may be worthwhile improving the usage explanations in the MAF text. The BoG will consider this suggestion further.14511 Florian Schreiner Chun Hui Suen

Overview of ISO/IEC 23000-7 CD Open Release MAF (1-pager)

14513 Florian Schreiner Chun Hui Suen

Proposed text to ISO/IEC 23000-7 CD Open Release MAF

Input:M14511 provides an overview of the OR MAF.

Actions:

92

The BoG will consider the overview an d then MDS will create an output document of the overview for the web site. One issue is whether the relationship to CC rights should be made explicit.Input:M14513 provides improved text for the CD.

Actions:The BoG used this text as a basis for work during the week. The BoG also considered rights issues brought out.

14503

Hélder Castro Pedro Carvalho Teresa Andrade Christian Timmerer Hermann Hellwagner

A DID model for Media Streaming MAF

Input:This input proposed a constrained DID – a model /profile – for use in the MS MAF. The model contains Descriptors which cater for each stakeholder in the DI delivery chain. A possible problem with referencing Digital Items was also identified.

Actions:This input is related to the BoG activities on a new improved, royalty free DI. The requirements of this application will be considered during those discussions. It is also envisaged that the input may have impact on the MS MAF work and will be discussed in that BoG.MPEG-7 Query Format (16h30 - 18h00)   MDS Room FIR

This was the first meeting in MDS of the MP7QF BoG.14365 Davide Rogai Paolo Nesi Pierfrancesco Bellini

Experience on using MPEG-21 File Format for nested and/or protected DIs

Input:This input considered some problems that were encountered using the MPEG-21 FF and DIDs with protected content requirements. One use case is where a piece of content has been protected once and then protected with a second technology.

Actions:There are various solutions involving layered protection in the ISO FF and then also through the use of MPEG-21 IPMP Components. No further action at this meeting. The authors of the input will try the layers of solutions and report back at a future meeting.

14351 Saar De Zutter Jan De Cock Rik Van de Walle

Conformance tests for DIDL documents - files

14356

Saar De Zutter Jan De Cock Rik Van de Walle on behalf of the Belgian National Body

BNB comments on ISO/IEC FCD 21000-14: Conformance Testing

14409

Saar De Zutter Jan De Cock Rik Van de Walle on behalf of the Belgian National Body

Preliminary BNB comments on ISO/IEC FCD 21000-8: Reference Software (2nd edition)

Input: 14351These are the XML files for testing DIDs

Actions:These should be attached to the Conformance FDIS

Input: 14356

93

Belgain NB comments on the conformance document. Recommends accepting the Study document an d the annex referencing, spacing, explanations in Annex A have incomplete sentences.

Actions:These changes should be incorporated into the Conformance FDIS.

Input: 14409Preliminary comments on the Reference software from Belgium.

Actions:These should be added to the study of the Reference software.

14462

Michael Eberhard Christian Timmerer Hermann Hellwagner

Update of gBSDtoBin and DIA Reference and Utility Software Modules

Input: This input updates the gBSDtoBin and DIA Ref/Utility software modules. Actions:Add these modules to the reference software, replacing older modules

14505 Eva Rodríguez Jaime Delgado

Contribution to MPEG-21 Reference Software: Validation Rules Checker for the REL MAM Profile

14401 Eva Rodríguez Jaime Delgado

Contribution to REL MAM Profile Conformance

Input: 14401This input suggests mechanisms for Conformance for the REL MAM profile. It suggests creating a subset of the REL rules for conformance and one new rule. Reference software checking the rules is available.

Actions:Add these modules to the reference software, replacing older modules

Input: 14505This software implements the rules specified in m14401.

Actions:Add these modules to the reference software (study document).

14399 Eva Rodríguez Jaime Delgado

Adding Integrity and authenticity to Event Reporting information

14400 Jaime Delgado Eva Rodríguez

Defect Report Proposal of ISO/IEC 21000-15

Input: 14399This input raises again the possibility of adding security to ER. It proposes using both MPEG and non-MPEG standards. The integrity is provided using Digital Signatures. For data encryption: XML encryption.

Actions:MDS agrees this is useful. However, modifications to the standard are only required if there is a use case for protecting ‘part’ of an ER.

94

Input: 14400This input raises problems with the inData element of ER. The standard is inconsistent between the text and schema.

Actions:To correct the schema in Annex A

14508 Eva Rodríguez Jaime Delgado Víctor Torres

Some issues on the generation and modification of Event Reports in the MPEG-21 Event Reporting

Input: This input raises issues with MPEG-21 ER. It suggests adding structure to Descriptor, a new child element to modification or introducing multiple ER report elements.

Actions:This would be a useful correction to the current ER specification as multiple ER report elements is important in various applications.14502

Daniel Oancea Pedro Carvalho Teresa Andrade Christian Timmerer Hermann Hellwagner

Defect Report on ISO/IEC 21000-15

Input: This input raises three scenarios for the use of ER: Terminal, Service monitoring, network related. It recommends removing the requirement for an ER Request. Also, highlights several inconsistencies in ER, and that the schema documents are not extensible.Actions:This would need a Corrigenda to be created on ER. MDS will issue a DCOR at this meeting. 14481 Hendry Takafumi Ueno

Some Editorial Update for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14482 HendryLate comment for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14483 Hendry Munchurl Kim

Contribution for MPEG-21 IPMP Components Base Profile Conformance

Input: 14481This input describes a series of editorial updates for IPMP Components Base profile. These address the comments of the Japanese National Body.Actions:MDS accepts the editorial changes and will issue these as part of the AMD/1

Input: 14482This is a late comment supporting editorial changes and was withdrawn.

Input: 14483This input provides the conformance for the IPMP Components Base Profile. It provides test sequences according to the Based Profile Restrictions. Each test sequence provides for testing of instances one by one.

Actions:This will be included in the Conformance FDIS and will be supported by an NB comment.

14557 Christian Timmerer Late Austrian NB comments on ISO/IEC 21000-7 Cor.1

95

(on behalf of ANB)

Input: This input notes that FGS has been removed from the FDIS of SVC and hence the descriptors must be adjusted in ISO/IEC 21000-7.

Actions:This will be accepted and the adjustments made.

Note on the Professional Archival MAFAt the San Jose meeting, the support for the Professional Archival MAF was discussed and the consensus was that at this stage there is not enough support for progression of this MAF on the standards track. At this stage the MAF part is on hold pending increased support for the activity.

The Break Out group reports for the meeting are given below:

MPEG-7 Query

96

97

98

REL/Open Release Discussions

BoG of REL and OR MAFApril 24th, 2007

MPEG-21 Part 5: REL profiles

DAC

14415 2007-04-17 2007-04-16 MPEG-21 MDSKisong YoonTaehyun KimHogab Kang

Interoperability between MPEG-21 REL DAC Profile and Other Standards

1. MPEG white paper, with three weeks editing period.

2. OutputRegulation a. with no occurrence, the behavior is to allow output the source content signal (or into any

possible output signal) – Satoshi b. with occurrence but no child elements, the behavior is to allow output/preserve the source

content signal – Taehyun c. with occurrence and child elements, the behavior is to allow output signal according to the

constraint in the child elements.

99

ORC

1. Dispose NB comments from Spain and Korea

2. For embedding: embed right for source content and enlarge or enhance right for target content

Use case Right for Source Right for Target

Embedding a source into a new target Embed AdaptEmbedding a source into an existing targetAdapt Adapt N/AAggregate Aggregate N/A

2. FPDAM/3

14479 2007-04-18 2007-04-16 MPEG-21 MDS

Taehyun KimJaime Delgado

Florian SchreinerChris Barlas

Editor's study of ISO/IEC 21000-5/PDAM3

14484 2007-04-18 2007-04-16 MPEG-21 MDSKisong YoonTaehyun KimHogab Kang

A Study on Use Cases of Derivative Works with MPEG-21 REL ORC Profile License

1. Derivative works with their derived licenses

MPEG-21 Part 8: Reference Software (for REL profile sections)

14505 2007-04-18 2007-04-16 MPEG-21 MDS Eva RodríguezJaime Delgado

Contribution to MPEG-21 Reference Software: Validation Rules Checker for the REL MAM Profile

1. Update the software plan and Part 8. 2. (informative) License creator is still missing for DAC. Need it for the next meeting.3. Update the software plan (version 6) to include modules for ORC

MPEG-21 Part 14: Conformance Testing (for REL profiles sections)14401 2007-04-18 2007-04-13 MPEG-21 MDS Eva Rodríguez

Jaime Delgado Contribution to REL MAM Profile Conformance

1. Add the rules for MAM to the spec. 2. DAC may need new rules and thus will be considered at the next meeting.

OR MAF1. DIDL input text

a. Incorporate the input text into text of 23000-7 FCD 2. Relationship with CC licenses

a. Why including an ORC license in an OR MAF content/package?i. To provide a mechanical means to enable and help users to manage (use,

adapt and distribute) Open Release content

100

b. Why including an identifier, name, or link to a CC license in an OR MAF content/package? Choose one of the following:

i. To indicate the intentions of the OR content creator as expressed in the CC license

ii. To provide CC license information as metadata for the legal notification purpose (depending on positive feedback from CC)

c. When both a CC license and an ORC license are present in a same OR package, i. the CC license is for information only, and the ORC license is for usage

managementii.

d. Can OR package only mention CC license names but not CC license links or content, in order to avoid any legal issues?

3. CC comments from Mike.

The example includes...<CopyrightString> Creative Commons (CC) License: Attribution Non-commercial No Derivatives (by-nc-nd) </CopyrightString></Creation> <RelatedMaterial> <MaterialType> <Name>Licensing Information Page</Name> </MaterialType> <MediaLocator> <MediaUri>http://creativecommons.org/licenses/by-nd-nc/1.0/</MediaUri> </MediaLocator></RelatedMaterial>

The content of <CopyrightString> is presumably a notice for humans.Is the value of <Name>, that is "Licensing Information Page", from a controlled vocabulary? Or is that just informational for humans as well?

In the Open Release MAF we use at the moment the MPEG-7 "RelatedMaterial" element to provide related information such as the link to a related CC license. The "Name" in Materialtype is only a piece of information for humans. To address your concern, we will consider how to provide some information to say that the related material is in fact a reference to the CC license.

Regarding the "Rights Expression Language, AMENDMENT 1: MPEG-21 REL profiles" document, which the "Open Release MAF" document says "defines rights and conditions for modelling creative-commons like licenses."

It looks like the right primitives are present, though I'm not sure I understand how each is expressed. Take "Figure1 - m3x:governedAdapt Right" which is described as "any principal is granted the right to play a movie clip, and the right to adapt the clip together with the same license."

I don't see where in the example "with the same license" is expressed. Is this implicit?I also do not see any means for explicitly identifying the license used.Even if the rights associated with a CC license are accurately described the specific license should be identified with a license URI.

Regarding the question on "with the same license", it means the same license which original content has. More precisely, the right "m3x:governedAdapt" needs to make another license which is same as the original one the right is part of when it is exercised. So currently it does not have to have a specific identifier. However, if it is needed, an identifier can be specified for the original license.

101

4. More questions to CCa. Merging two conflicting sharelike licenses (e.g., commercial and non-commercial).

Output Documents1. DoC of 21000-5 PDAM/3 -- Done2. Text of 21000-5 FPDAM/3, two weeks editing period – Taehyun and Jaime3. Output on DAC interoperability with other rights information standards – Taehyun, Xin,

Jaime, Satoshi4. REL/RDD reference software development plan v6 – Florian and Xin5. DoC of 23000-7 CD – Florian6. Text of 23000-7 FCD, with four weeks editing period – Florian

7. Contribution to MPEG-21 Parts 8 and 14 – Jaime and Xin

3.0 MDS Output Documents and Resolutions – San Jose 80th Meeting

The MDS subgroup recommends approval of the following documents

MPEG-7

No. Title TBP Available15938-5 Multimedia Description Schemes

9129 DoC on ISO/IEC PDAM/1 15938-5 Improvements to Geographic Descriptor

07/04/27

9100 ISO/IEC FPDAM/1 15938-5 Improvements to Geographic Descriptor

07/04/27

No. Title TBP Available15938-7 Conformance testing

9130 DoC on ISO/IEC PDAM/1 15938-7 Improvements to Geographic Descriptor Conformance

07/04/27

No. Title TBP Available15938-10 Schema definition

9102 Schema Files for MPEG-7 07/04/27

1.1.3. The MDS subgroup notes that the document NXXXX is a first version of an ongoing working document containing the ‘electronic’ versions of schemas for the current MPEG-7 parts at IS/FDIS. The MDS subgroup requests that the versions of the schemas be updated on the ITTF WWW site at a similar URL to the equivalent MPEG-21 schemas.

1.1.4. The MDS subgroup also requests that users of the schemas who choose to create reduced or profiled schemas input these to MPEG so MPEG might understand usage of the MPEG-7 descriptors. Further details are provided with the schema files.

1.1.5. The MDS subgroup recommends appointing Robert O'Callaghan and Akio Yamada as the editors of ISO/IEC 15938-10:2005/COR 1 and thanks them

102

for taking responsibility for that project.

No. Title TBP Available15938-12 MPEG-7 Query Format

9103 ISO/IEC 15938-12 CD MPEG-7 Query Format 07/05/259104 Technologies Under Consideration for MPEG-7 Query Format 07/04/27

1.1.6. The MDS subgroup recommends appointing Kyoungro Yoon, Mario Doeller, Matthias Gruhne, Ruben Tous, Masanori Sano, Miran Choi, Tae-Beom Lim, Jongseol James Lee and Hee-Cheol Seo as the editors of ISO/IEC 15938-12 MPEG-7 Query Format and thanks them for taking responsibility for that project.

MPEG-21

No. Title TBP Available21000-4 IPMP Components

9105 DoC of ISO/IEC 21000-4 FPDAM/1 IPMP Components Base Profile

07/04/27

9106 Text of ISO/IEC 21000-4 FDAM/1 IPMP Components Base Profile 07/04/27

1.1.3. The MDS subgroup thanks the National Body of Japan for their useful comments on ISO/IEC PDAM/1 21000-4.

No. Title TBP Available21000-5 Rights Expression Language

9107 DoC of ISO/IEC 21000-5 PDAM/3 ORC (Open Release Content) Profile

07/04/27

9108 ISO/IEC 21000-5 FPDAM/3 ORC (Open Release Content) Profile 07/05/25

9109 Interoperability between MPEG-21 REL DAC Profile and other Rights Information Standards

07/05/18

9110 REL/RDD Reference Software Development Plan v.6 07/04/27

1.1.4. The MDS subgroup thanks the National Bodies of Korea, Japan and Spain for their useful comments on ISO/IEC PDAM/3 21000-5.

No. Title TBP Available21000-7 Digital Item Adaptation

9111 Disposition of Comments on ISO/IEC 21000-7:2004/DCOR 1 07/04/27

9112Text of ISO/IEC 21000-7:2004/COR 1 MPEG-21 Digital Item Adaptation

07/05/25

9113 Text of ISO/IEC 21000-7 FDIS Second edition 07/05/25

103

The MDS subgroup recommends appointing Christian Timmerer (Klagenfurt University), Sylvain Devillers (France Telecom), and Michael Ransburg (Klagenfurt University) as the editors of ISO/IEC 21000-7 2nd edition and thanks them for taking responsibility for that project.

The MDS subgroup recommends appointing Christian Timmerer (Klagenfurt University) as the editor of ISO/IEC 21000-7:2004/COR and thanks him for taking responsibility for that project.

1.1.3. The MDS subgroup thanks the National Bodies of Austrian and France for their useful comments on ISO/IEC DCOR/1 21000-7.

No. Title TBP Available21000-8 Reference Software

9114 Preliminary DoC of preliminary comments of ISO/IEC 21000-8 FCD Reference Software

07/04/27

9115 Study text of ISO/IEC 21000-8 FCD Reference Software 07/04/27

No. Title TBP Available21000-14 Conformance

9116 DoC of ISO/IEC 21000-14 Conformance 07/04/279117 Text of ISO/IEC FDIS 21000-14 Conformance 07/05/25

1.1.4. The MDS subgroup thanks the National Bodies of Australia, Austria, Belgium, France, Korea, Spain and the US for their useful comments on ISO/IEC FCD 21000-14.

No. Title TBP Available21000-15 Event reporting

9118 ISO/IEC 21000-15:2006/DCOR1 MPEG-21 Event Reporting 07/05/21

The MDS subgroup recommends appointing Christian Timmerer (Klagenfurt University) and Jaime Delgado (DMAG) as the editors of ISO/IEC 21000-15:2006/DCOR1 and thanks them for taking responsibility for that project.

No. Title TBP Available21000-18 Digital Item Streaming

9119 DoC of ISO/IEC 21000-18/PDAM 1 07/04/279120 ISO/IEC 21000-18/FPDAM/1 Simple fragmentation rule 07/06/08

WG11 thanks the International Confederation of Societies of Authors and Composers (CISAC) for its current role in serving as the Registration Authority (RA) for ISO/IEC 21000-3. WG11 requires the services of an RA for ISO/IEC 21000-18 and has determined the requirements to be compatible with those of the RA for ISO/IEC 21000-3 and has received a letter of in principle agreement from CISAC to serve as RA for ISO/IEC 21000-18. WG11 therefore requests the SC29 secretariat to issue the ballot asking for CISAAC to be appointed Registration Authority for ISO/IEC 21000-

104

18.

MPEG-A

No. Title TBP Available23000-2 Music Player Application Format

9121 DoC of ISO/IEC 23000-2 FCD Music Player Application Format 2nd Edition

07/04/27

9122 Text of ISO/IEC 23000-2 FDIS Music Player Application Format 2nd Edition

07/05/25

1.1.5. The MDS subgroup thanks the National Bodies of Germany, Japan and the UK for their useful comments on ISO/IEC FCD 23000-2.

No. Title TBP Available23000-5 Media Streaming MAF

9123 DoC on ISO/IEC CD 23000-5 Media Streaming Player 07/04/279124 ISO/IEC FCD 23000-5 Media Streaming Player 07/05/25

1.1.3. The MDS subgroup thanks the National Bodies of Austria, Korea and the UK for their useful comments on ISO/IEC CD 23000-5.

No. Title TBP Available23000-7 Open Release Application Format

9125 DoC of ISO/IEC 23000-7 CD Open release MAF 07/04/279126 ISO/IEC 23000-7 FCD Open release MAF 07/05/25

1.1.3. The MDS subgroup thanks the National Body of Spain for their useful comments on ISO/IEC CD 23000-7.

No. Title TBP Available23001-5 Bitstream Syntax Description Language

9127 Text of ISO/IEC 23001-5 FDIS Bitstream Syntax Description Language

07/07/01

The MDS subgroup recommends appointing Sylvain Devillers and Joe Thomas-Kerr as the editors of ISO/IEC 23001-5 and thanks them for taking responsibility for that project.

9128 AHG on MPEG-7 Query FormatMandate: To address the following issues:

1. Complete editing of the MPEG-7 Query Format CD2. Consider improvements to the CD and TuC documents3. Continue discussions on Server Selection and Capabilities4. Study the functionality and syntax of operators

105

Chairman: Kyoungro Yoon (yoonk *at* konkuk.ac.kr)Mario Doeller (Mario.doeller_*at*_uni_passau.de)

Duration: Until 81st Meeting Meetings AHG meeting will be held on the weekend prior to 81st meeting. Other business

will be conducted by e-mail or telephone conference.Reflector: cbsearch@yahoogroups.comSubscribe: To subscribe send email to cbsearch-subscribe@yahoogroups.com

106

4.0 MDS Final Schedule – San Jose 80th Meeting

  MPEG MDS Chair: Ian S Burnett

  MPEG-7, MPEG-21, MAF v3.0

Number Source Title

     Monday Morning (9h00-13h00)    MPEG Plenary   Plenary room

     Monday Afternoon (13h30-20h00)    

     Kick-off of MPEG MDS activities (13h30-14h00)   MDS Room FIR

 

Agenda, Goals and Issues for the Week for MDS Group Ian S Burnett

     Review of AHG resolutions, CE results and action points (13h30-14h20)   MDS Room FIR

14277Gerrard Drury Peder Drege AHG on MPEG-21 DIS

14278

Filippo Chiariglione Christian Timmerer Thomas Skjolberg AHG on the Media Streaming MAF demo for the MAF-AE

14279

Stefan Kraegeloh Filippo Chiariglione Noboru Harada AHG on MDS MAFs Under Development

14280

Wo Chang Kyoungro Yoon Mario Doeller AHG on MPEG-7 Query Format

14537Masanori Sano Hideki Sumiyoshi Nobuyuki Yagi

Test report of CEs on MP7QF

14538Masanori Sano Hideki Sumiyoshi Nobuyuki Yagi

Test report of CE on specification of the request of the Output

14539Masanori Sano Hideki Sumiyoshi Nobuyuki Yagi

Test report of CE on Query operation based on text description

14543 Ruben Tous Jaime Delgado DMAG CE Report for CEs on MPEG-7 Query Format

14524 Saar De ZutterReview of Core Experiment on query operation based on text description

14330

Thomas Skjølberg Peder Drege Joseph Thomas-Kerr Gerrard Drury

Report of CE on DIS TuC

14458

Ingo Kofler Christian Timmerer Hermann Hellwagner on behalf of Austrian NB

Austrian NB comments on ISO/IEC 21000-7 Cor.1

14459Michael Eberhard Christian Timmerer

Austrian NB comments on ISO/IEC 21000-8 FCD

Hermann Hellwagner on behalf of Austrian NB

14460

Christian Timmerer Hermann Hellwagner

Austrian NB comments on ISO/IEC CD XXXXX Media Streaming MAF Protocols

14461

Christian Timmerer Michael Ransburg Hermann Hellwagner

Austrian NB comments on ISO/IEC 23000-5 CD

Define BoGs and Mandates (14h20-14h30)   MDS Room FIR    BoG1 = San Carlos       MPEG-7 QF Zinfandel  OR MAF Tues am   Prof Archival    DID Mon 4.30-6pm  MS MAF Wed 4pm  REL Profiles Tues am DIA Futures (14h00 - 14h30)   MDS Room FIR

14318 Sylvain Devillers Editors' input to draft text of 23001-5 (MPEG-B BSDL)

14341

Christian Timmerer Sylvain Devillers Michael Ransburg

Editor's input on Draft MPEG-21 DIA 2nd edition

REL (14h30 - 15h30)   MDS Room FIR     

14415Kisong Yoon Taehyun Kim Hogab Kang

Interoperability between MPEG-21 REL DAC Profile and Other Standards

14479

Taehyun Kim Jaime Delgado Florian Schreiner Chris Barlas

Editor's study of ISO/IEC 21000-5/PDAM3

14484Kisong Yoon Taehyun Kim Hogab Kang

A Study on Use Cases of Derivative Works with MPEG-21 REL ORC Profile License

 Open Release MAF/MS MAF (15h30 - 16h30)   MDS Room FIR

14507Eva Rodríguez Jaime Delgado

Contribution to the current version of the Open Release MAF

14511Florian Schreiner Chun Hui Suen

Overview of ISO/IEC 23000-7 CD Open Release MAF (1-pager)

14513Florian Schreiner Chun Hui Suen

Proposed text to ISO/IEC 23000-7 CD Open Release MAF

14442Hyon-Gon Choo Filippo Chiariglione

Proposed text of ISO/IEC 23000-5 FCD Media Streaming MAF

14503

Hélder Castro Pedro Carvalho Teresa Andrade Christian Timmerer Hermann Hellwagner

A DID model for Media Streaming MAF

     MPEG-7 Query Format (16h30 - 18h00)   MDS Room FIR     

     Tuesday Morning

   

108

(9h00-13h00)     

     MDS/Reqts issues (09h00 - 11h00)   Reqts     

14500 Sylvain Devillers Use of MPEG URN for identifying profiles and levels

14532 Gerrard DruryContribution on URI assets and Requirements and Structure of URNs

14475

Giovanni Cordara (on behalf of the ITNB) Italian NB proposal to revisit MPEG-21 DID

14420

Hee-Cheol Seo Miran Choi Hyunki Kim Myung-Gil Jang Soojong Lim Jeong Heo Kyoungro Yoon CE Report for Query Expression of MPEG-7 Query Format

14421

Hee-Cheol Seo Miran Choi Hyunki Kim Myung-Gil Jang Soojong Lim Jeong Heo Kyoungro Yoon

Revision of Proposed Input Query Format for MPEG-7 Query Format

14449 Doeller Gruhne Wolf MP7QF CE Test Report 

MDS/Systems DID (11h00 - 12h00)   MDS

14365Davide Rogai Paolo Nesi Pierfrancesco Bellini

Experience on using MPEG-21 File Format for nested and/or protected DIs

          Tuesday Afternoon (14h00-18h00)         MAFs (14h00 -18h00)    

14430 Tilman Liebchen Comments on Professional Archival MAF Requirements

14411Noboru Harada Takehiro Moriya Yutaka Kamamoto

Proposed text to WD of Professional Archical MAF

14352James Orwell James Annesley Contribution to the Basic Video Surveillance MAF

14486Houari Sabirrin Jeongyeon Lim Munchurl Kim A Proposal for Basic Video Surveillance Application Format

14477Hendry Houari Sabirin Munchurl Kim

Updated Proposal for Protected Musical Slide Show MAF with IPMP

14478Hendry Houari Sabirin Munchurl Kim Updated Proposal for Protected Photo Player MAF with IPMP

14423

Kwangcheol Choi Sung-Moon Chun Jaedo Kwak Seungheon Yang Ji-Sang Yoo Si-Hun Sung Seong-Cheol Han Requirements for Stereoscopic MAF

14424

Jaedo Kwak Si-Hun Sung Sung-Moon Chun JinWoong Kim Namho Hur Whitepaper of Stereoscopic Project

 

14367

Paolo Nesi Pierfrancesco Bellini Davide Rogai Kia Ng (University of Leeds)

Proposal for a MAF on Cross-Media Interactive Presentation: Overview and Application Scenarios

14368Paolo Nesi Pierfrancesco Bellini Davide Rogai

Proposal for a MAF on Cross-Media Interactive Presentation: Requirements

14369Davide Rogai Pierfrancesco Bellini Paolo Nesi

Proposal for a MAF on Cross-Media Interactive Presentation: Relationships with other MAFs

      

     Wednesday Morning

   

109

(09h00-13h00)     MPEG Plenary (9h00-11h00)   Plenary roomConformance /Ref s/w (11h00-12h00)   MDS Room FIR

14351Saar De Zutter Jan De Cock Rik Van de Walle

Conformance tests for DIDL documents - files

14356

Saar De Zutter Jan De Cock Rik Van de Walle on behalf of the Belgian National Body

BNB comments on ISO/IEC FCD 21000-14: Conformance Testing

14409

Saar De Zutter Jan De Cock Rik Van de Walle on behalf of the Belgian National Body

Preliminary BNB comments on ISO/IEC FCD 21000-8: Reference Software (2nd edition)

14462

Michael Eberhard Christian Timmerer Hermann Hellwagner

Update of gBSDtoBin and DIA Reference and Utility Software Modules

14505Eva Rodríguez Jaime Delgado

Contribution to MPEG-21 Reference Software: Validation Rules Checker for the REL MAM Profile

14401Eva Rodríguez Jaime Delgado

Contribution to REL MAM Profile Conformance

     MPEG-7 Discussions - Schema (12h00-12h30)   MDS Room FIR     

     Wednesday Afternoon (14h00-17h45)    ER (14h00 - 15h00)   MDS Room FIR

14399Eva Rodríguez Jaime Delgado

Adding Integrity and authenticity to Event Reporting information

14400Jaime Delgado Eva Rodríguez

Defect Report Proposal of ISO/IEC 21000-15

14502

Daniel Oancea Pedro Carvalho Teresa Andrade Christian Timmerer Hermann Hellwagner

Defect Report on ISO/IEC 21000-15

14508Eva Rodríguez Jaime Delgado Víctor Torres

Some issues on the generation and modification of Event Reports in the MPEG-21 Event Reporting

 IPMP Components (15h00 - 16h00)   MDS Room FIR

14481 Hendry Takafumi UenoSome Editorial Update for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14482 HendryLate comment for ISO/IEC 21000-4/FPDAM1 MPEG-21 IPMP Components Base Profile

14483 Hendry Munchurl KimContribution for MPEG-21 IPMP Components Base Profile Conformance

     MPEG-21 & LASeR (16h00 - 17h00)   SYSTEMS

14418

Jihun Cha YeSun Joung Young-Kwon Lim KyungAe Moon Ideas on MPEG-21 and LASeR

Thursday Morning (9h00-12h30)    

110

     Breakout Issues         MPEG-7 QF(11h00-12h00)   Reqts joint with JPEGThursday Afternoon (14h00-19h00)    MPEG-7 Query Discussions (14h00 - 15h00)   MDS Room FIRDID discussions(15h00-16h00)   ReqtsPlenary MDS and Reports of BoG (16h00 - 18h00)   MDS Room FIR          Further review of Output documents, AHGs, CEs, DoC, Std (18h00+++)   MDS Room FIRFriday Morning (09h00-13h00)    

Wrapping up (09h00 - 13h00)   MDS Room FIR

 

Approval of resolutions, AHGs and Output documents  

     Friday Afternoon (14h00-21h00)    MPEG Plenary   Plenary roomContact: Ian S Burnett    

x    

111

Annex H – Video report

Source: Jens-Rainer Ohm, Gary J. Sullivan (Video), Miroslaw Z. Bober (MPEG-7 Visual)

44 MPEG-4 Visual Simple Profile Level 6

The specification text and conformance part related to the new level 6 (720p resolution) of MPEG-4 Visual Simple Profile have progressed as expected. Comments made by NBs ad been mostly of editorial nature.

Documents reviewed:14383 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-2:2004/PDAM 4

14386 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/PDAM 28

Documents approved:No. Title TBP Available

14496-2 Visual8948 Disposition of Comments on ISO/IEC 14496-2:2004/PDAM4 No 07/04/278949 Text of ISO/IEC 14496-2:2004/FPDAM4 Simple Profile Level 6 No 07/04/278952 Disposition of Comments on ISO/IEC 14496-4:2004/PDAM28 No 07/04/278953 Text of ISO/IEC 14496-4:2004/FPDAM28 Visual Simple Profile

Level 6 Conformance TestingNo 07/04/27

45 MPEG-4 Video Conformance Corrigenda

Errors in MPEG-4 Video conformance bitstreams (incorrect signaling of low delay mode) were reported in 14358. It was decided to go for a new corrigendum directly (no defect report before) because currently no other problems with conformance streams are envisaged. In the same context, an editorial error which occurred by the time when the 2004 edition was produced is corrected, where bitstreams relating to studio profile, FGS, ASP and new levels were by mistake attached to the new edition, even though their description is only included in Amd.1 and Amd.3 of the new edition. In fact, various studio profile and ASP stream would now have been missing in Amd.1, such that another corrigendum on that part became necessary.

Documents reviewed:14358 Yi-Shin Tung, Ja-Ling Wu Additional fixes on MPEG-4 video conformance

bitstreams

Documents approved:No. Title TBP Available

14496-4 Conformance testing8950 Text of ISO/IEC 14496-4:2004/DCOR4 No 07/04/278951 Text of ISO/IEC 14496-4:2004/Amd.1/DCOR2 No 07/06/29

112

46 MPEG-7 Visual

46.1 MPEG-7 Visual related work in San JoseThe MPEG-7 breakout group was active during the whole week. Input documents related to the Visual part in 15938-3 and Photo Player MAF (23000-3) are listed in the table below. All of these documents were reviewed and discussed.

14350

Weon-Geun Oh, Ju-Kyoung Jin, A-Young Cho, Jun-Woo Lee, Ik-Hwan Cho, Won-Keun Yang, Dong-Seok Jeong

Mathematical consideration on the degree of geometrical modification

14406 Sangki Kim, Hyobin Lee, Sangyoun Lee CE Report for VCE-5

14412 Weon-Geun Oh, Won-Keun Yang, Dong-Seok Jeong

Modified GST Based Descriptor for MPEG-7 VCE-6 Complex Condition

14436 Ryoma Oami CE report for VCE-3 on person identity-based photo indexing

14439 Kota Iwamoto, Ryoma Oami CE report for VCE-7 on video signature14440 Kota Iwamoto, Ryoma Oami Proposal of CE procedure for VCE-7

14470 Paul Brasnett, Miroslaw Bober Improved Image Identifier (VCE6)

14471 Paul Brasnett, Miroslaw Bober Modification of VCE6 Experimental Conditions

14472 Paul Brasnett, Miroslaw Bober VCE7 Experimental Conditions

14523A-Young Cho, Ik-Hwan Cho, Jun-Woo Lee, Weon-Geun Oh, Dong-Seok Jeong

New Visual Identifier for MPEG-7 VCE-6 Basic Condition

Summary of key work items: Review of the Core Experiment results and future planning

VCE-3 – Face-bases Annotation VCE-5 - Evaluation of MPEG-7 Face Recognition Technology on IR

Images VCE-6 - Image Signatures VCE-7 – Video Signatures

Photo Player MAF S/W development – second version Review of the Protected Photo Player proposal Study Text of ISO/IEC 23000-3/PDAM1 Reference Software for Photo Player

MAF Editorial work, Maintenance and Software development Joint meeting with JPEG-search

Results of Core Experiments:

Much of the time during the week was spent discussing the core experiment on Visual Identifiers (VCE-6). An improved version of the current XM algorithm was presented, based on the Trace Transform (see M14470). There was a competing contribution (M14523), using the “concentric circle-based visual identifier", which showed inferior performance to M14470, despite exceeding that of the previous XM version. Consequentially, the proposed modifications of M14470 were adopted. The experimental conditions were tightened for the continuing CE, in order that differences between the algorithms may become more apparent, by the next meeting. In revising the experimental conditions, account was taken of two other proposals (M14552 & M14471). The image database used in the CE for independence testing has itself been found to contain

113

several duplicated images. Significant time and effort was dedicated to agreeing which pairs of images are modified copies of one another and which are independent (i.e., different) images. Final list will be agreed by consensus on the reflector.

There was also a contribution (M14412) to the other part ("Complex Condition") of VCE-6, but this was deemed to not yet have performance sufficient for adoption into the XM.

There was one contribution each for VCE3 on person-identity-based photo clustering (M14436) and VCE-5 on IR-sensor-based face recognition (M14406). Both experiments will continue. The former has been hampered by the lack of source code for the (prospective) reference method, from Samsung. In the latter, a key milestone in the coming period will be the distribution of an IR face-image database for other participants to share.

Three contributions were made to VCE-7 on the Video Identifier (M14439 , M14440, and M14472). Video sequences used previously in MPEG-7 visual core experiments were shared amongst the participants for use in VCE-7; however, sufficient test material has not yet been accumulated for the independence test (in which the recall-bias will be set to achieve a pre-determined false positive rate). The experimental conditions were amended in response to the inputs.

New Amendment:

The working draft of a new amendment to 15938-3 (defining Visual identifiers for different purposes) is planned for the July meeting.

Editorial work, Maintenance and Software:

COR1 of 15938-3:2002/Amd.2 was produced (related to perceptual 3D shape) COR1 of 15938-6:2003/Amd.1 was produced (related to color temperature) FDAMs of software and conformance related to the Perceptual 3D Shape descriptor were

produced Study of PDAM1 of 23000-3 was produced (new stabilized version of reference software)

46.2 Output documents related to MPEG-7 Visual

No. Title TBP Available15938-3 Visual

8969 Text of ISO/IEC 15938-3:2002/Amd.2:2006/Cor.1 (Perceptual 3D Shape)

No 07/04/27

8970 MPEG-7 Visual XM Document version 30.0 No 07/04/278971 Description of Core Experiments for MPEG-7 New Visual

ExtensionsNo 07/04/27

15938-6 Reference Software 8972 Disposition of Comments on ISO/IEC 15938-6:2003/

Amd.1:2006/DCOR 1No 07/04/27

8973 Text of ISO/IEC 15938-6:2003/Amd.1:2006/Cor.1 (Color Temperature)

No 07/04/27

8974 Disposition of Comments on ISO/IEC 15938-6:2003/FPDAM2 No 07/04/278975 Text of ISO/IEC 15938-6:2003/FDAM2 (Perceptual 3D Shape) No 07/04/27

15938-7 Conformance testing8976 Disposition of Comments on ISO/IEC 15938-7:2003/FPDAM3 No 07/04/278977 Text of ISO/IEC 15938-7:2003/FDAM3 (Perceptual 3D Shape) No 07/04/27

114

46.3 Output documents related to MPEG-A Photo Player MAF

No. Title TBP Available23000-3 Photo Player Application Format

8978 Study Text of ISO/IEC 23000-3/PDAM1 Reference Software for Photo Player MAF

No 07/05/25

47 Misceallanea

14468Nicola Adami, Riccardo Leonardi, Pierangelo Migliorati, Claudia Tonoli

Performance of a Distributed Video Codec in Presence of Transmission Errors

The contribution reports that currently, Distributed Video Coding (DVC) approaches significantly lag behind the conventional codecs. It is however claimed that DVC would have good properties in terms of error resilience. For the experiments, a scenario is used where key frames are encoded by AVC, and Wyner-Ziv coded frames are interpolated from them (similar as B frames, but without motion information). Good performance is found in case where the “side information” (key frames) is undistorted. In case where the side information (key frames) is distorted, intra error concealment is used in addition. It is shown that the Wyner-Ziv coded frames are quite robust, while keyframes show significant dropping. Rate for Wyner-Ziv coded frames is significantly higher than for key frames (which would be opposite in case of B frames). The contributors are noted that it would be interesting to compare this against a version where the overhead rate caused by Wyner-Ziv coding is rather used for error protection, or also unequal error protection as would be possible when using SVC with hierarchical B frames.

48 23002 MPEG-C Video Technologies

According to the current policy, any software and conformance supplement would be included in the respective part of MPEG-C, and not concentrated in a dedicated part as it had been the case for previous standards. Therefore, a resolution was issued recommending ITTF makes the reference software and conformance testing bitstreams of ISO/IEC 23002-x freely available on the ITTF website.

48.1 23002-1Again, an improved version of the software for IDCT conformance testing was provided for the FPDAM1 text. It not only contains the software which makes it possible to perform the precision test as described in the standard, but also allows to explore the performance of an IDCT in a full test bed, currently supporting MPEG-2 part 2 and MPEG-4 part 2.

No. Title TBP Available23002-1 Accuracy specification for implementation of integer-output IDCT

8980 Disposition of Comments on ISO/IEC 23002-1/PDAM1 No 07/04/278981 Text of ISO/IEC 23002-1/FPDAM1 Software for Integer IDCT

Accuracy TestingNo 07/05/31

48.2 23002-2 Fixed-point DCT/IDCT

The progression to FCD happened very smoothly, without any major conflicts.115

- Substantial input was provided by the editors for improving the overall quality of the text.- Based on results from CE, one small change in the algorithm was made (saving two shifts

without penalizing performance, see ISG report for more details)- It was decided to perform row-transform first, which has no impact on complexity or

performance, but is more consistent with other DCT/IDCT algorithms in the market, such that it may inflict less drift effect in cases where a different DCT/IDCT is used at the other end

- More investigations were made on the problem with the quarter-pel motion interpolation filter in MPEG-4 part 2, which imposes more critical drift when additionally different transforms are used at encoder and decoder. It has been verified that the DCT/IDCT algorithm of 23002-2 is in fact more resistant against this phenomenon

- It was planned to include a software implementation of the algorithm in the standard, and such software was added into the FCD.

- The word “implementation” was removed from the title of the draft standard, to avoid the impression that the standard requires a particular method of implementing the design (when in fact it prescribes only the result to be obtained by an implementation).

Documents reviewed:

14310 Yuriy A. Reznik, Gary Sullivan, Arianne T. Hinds Study Text of ISO/IEC 23002 CD (editors input)

14311 Yuriy Reznik Study Text of ISO/IEC 23002-1/PDAM1 (editors input)

14346 Yuriy Reznik Updated 23002-1 IDCT precision testbed14347 Yuriy Reznik, Arianne Hinds Updated H.263-based IDCT testbed14348 Arianne T. Hinds Updated MPEG-4 IDCT Testbed14359 Yi-Shin Tung, Ja-Ling Wu Consider row-transform-first IDCT in 23002-214379 Arianne T. Hinds Updated T.83 testbed for IDCTs14380 Zhibo Ni Updated MPEG-2 IDCT Testbed14403 Arianne T. Hinds Updated TM5 MPEG-2 Testbed

14469 Honggang Qi, Wen Gao, Debin Zhao, Siwei Ma Crosscheck for IDCT CD

14485 Zhibo Ni, Lu Yu IDCT Core Experiment Results

14506 Yuriy Reznik Summary of core experiments on fixed point IDCT/DCT

14509 Yuriy Reznik Cross-check of IDCT conformance tests14531 Arianne T. Hinds Fixed-Point IDCT Conformance Tests

14544 Zhibo Ni, Lu Yu On the Problem of Quarter Pixel Motion Compensation

Output Documents:

No. Title TBP Available23002-2 Fixed-point 8x8 IDCT and DCT

8982 Disposition of Comments on ISO/IEC CD 23002-2 No 07/04/278983 Text of ISO/IEC FCD 23002-2 Fixed-point 8x8 IDCT and DCT No 07/05/04

48.3 23001-4 and 23002-4 Reconfigurable Video Coding (RVC)

(High-level summary, for details on particular documents see ISG report)

MPEG-B related CE

14446MPEG-B Proposed Text of RVC CE

Notes reorganization of CE structure stated

116

Move CE 1.1 CE 2 (implementation)Recommendation Breakout meeting for each parties

14447MPEG-B Study on RVC Framework and Its Requirements

Notes Need to evaluate CE results with RVC requirements identified.

14445MPEG-B Core Experiment Result on CDDL

Notes Compression results are given

14340

MPEG-B Compression of the RVC DDL Decoder Description with BiM (results of Core Experiment 1.3 in RVC)

Notes Compression results are given

14473

MPEG-B Extension to support non-MPEG standards (ICT/ZJU) (Results of CE 1.6)

Notes Some modifications should be made on the design of syntax parsing

Restructuring of CE on MPEG-B part (CE 1) is done. Common ground of understanding and conducting CE is needed for better evaluation and

convergence between tools.

MPEG-C related CE (discussed on Tuesday)

14301 MPEG-C RVC Functional Units naming process proposal

14375 MPEG-C Conformance test tools of RVC functional units

14374MPEG-C Functional units of inter-prediction under reasonable system

partition for RVC framework

14416 MPEG-C Implementation of B frame support in RVC CAL Model

14454MPEG-C Implementation of multiple reference frame support in RVC CAL

model

14448 MPEG-C Proposed text of the RVC FUs for MPEG-4 AVC (Results of CE 2.2)

14457MPEG-C A scheme for implementing MPEG-4 SP codec in the RVC

framework

14480MPEG-C Implementation of MPEG-4 AVC Deblocking Filter in RVC

CAL model

14490MPEG-C Reconfigurability potential of the MPEG-4 SP decoder (results of

CE 1.1)

M14301: will be adopted in VTL WD. Further CEs will be continued. M14375: will be adopted in VTL conformance WD. Editor: Kris M14374: workplan will be updated for further FU implementation M14416: Bug fixing will be done in RSM implementation. M14454: will be adopted in VTL WD. M14448: FU textual description will be adopted in VTL WD & RSM. FU naming has to

follow the new naming rule M14457: For information. M14490: A figure should be added in the WD. The work will continue as CE 1.

117

M14480: FU textual description will be adopted in VTL WD & RSM.

Exploration Experiments (EE) related

14474MPEG-B Exploration experiments of AVS decoder description in RVC

framework

M14474: Will continue EE. EE should look at standardized token specification. Comparison with FUs.

Other issues14510 MPEG-C Proposal for adding ISO/IEC 23002-2 in RVC tool library

14463 MPEG-C Evolutions of RVC so as to handle SVC decoding

M14510: will be integrated into VTL WD. FU implementation will continue till the next meeting.

M14463: welcomes the contribution and expects more development by the next meeting.

Regarding the work plan (which was very ambitious), in particular the completeness of the tool library 23002-4 has not yet reached the expected status. Currently, it is estimated that only 20-30% of all MPEG video coding tools are fully described and implemented. It is of course useful to concentrate the work on the most relevant standards, however the tools of AVC baseline had not been finished before the San Jose meeting as originally planned; MPEG-4 Simple Profile (with fixed parser as FU) will be fully available next meeting (this is the minimum that should go into the first version of toolbox, or which may be added as future amendments). From the new work plan, it is expected that all of the most relevant profiles of MPEG-2, MPEG-4 Visual and MPEG-4 AVC will not be fully implemented before January 2008. Therefore, the time for producing the CD was decided to be delayed until July. In the case of 23001-4, one key issue is still the way how parsers can be constructed, for which two different solutions are currently on the table:

- via description of bitstream, possibly BSDL- as CAL-based FU(s)

In particular from the latter possibility, it is still necessary to clarify which parts of the parser will go into MPEG-B and MPEG-C. Further evaluation based on CE results will be necessary to find out which is the best solution. Evaluation criteria on this are agreed (see report of ISG for more details).

Liaison with AVS

14541 MPEG-B/MPEG-C

Liaison Statement to MPEG on RVC

- AVS has provided specification & software of their standard (as necessary for the current EE) to MPEG

- Even though there is no necessity for a “joint standard” on RVC, AVS representative(s) are highly welcome (as liaison) to participate in RVC development, in particular for the possibility of using 23001-4 with non-MPEG toolboxes.

- It is in MPEG’s own interest that the framework is generically applicable to non-MPEG standards

- Clear distinction between MPEG and non-MPEG toolboxes is necessary- A registration mechanism for non-MPEG toolboxes will be needed- To reflect the outcome of this discussion, the following wording is included in the RVC

project description: “The project is about developing a full collection of individual coding tools organized in the video tool library and a generic framework that can be used to make an

118

implementation of any MPEG video coding standard and additionally is capable of supporting the implementation of video coding standards from other organizations with which a collaboration can be established.”

- As part of this project, an identification mechanism will be developed whereby MPEG video coding tools will be identified by MPEG and video coding tools from other organizations can be identified via a registration authority.

- A new version of the requirements document will be edited, including a statement like this as well.

Documents reviewed:

14301Christophe Lucarz, Marco Mattavelli, Andrew Kinane, Sunyoung Lee, Sinwook Lee

RVC Functional Units naming process proposal

14340 Christophe Lucarz, Marco Mattavelli

Compression of the RVC DDL Decoder Description with BiM (results of Core Experiment 1.3 in RVC)

14374 Gwo Giun Lee, He-Yuan Lin, Ming-Jiun Wang

Functional units of inter-prediction under reasonable system partition for RVC framework

14375 Gwo Giun Lee, He-Yuan Lin, Ming-Jiun Wang Conformance test tools of RVC functional units

14416 Jar-Sheng Chen, Chun-Jen Tsai

Implementation of B frame support in RVC CAL Model

14445 Giseok Son, Sinwook Lee, Euee S. Jang Core Experiment Result on CDDL

14446 Hyungyu Kim, Euee S. Jang Proposed Text of RVC CE

14447 Jaebum Jun, Sunyoung Lee, Euee S. Jang Study on RVC Framework and Its Requirements

14448 Yoshihisa Yamada, Kenji Otoi, Kohtaro Asai

Proposed text of the RVC FUs for MPEG-4 AVC (Results of CE 2.2)

14454 Christophe Lucarz, Marco Mattavelli

Implementation of multiple reference frame support in RVC CAL model

14457

Ghislain Roquier, Maxime Pelcat, Mickaël RauletMatthieu Wipliez, Jean-François Nezan, Olivier Déforges

A scheme for implementing MPEG-4 SP codec in the RVC framework

14463

Maxime Pelcat, Médéric Blestel, Mickaël Raulet, Jean-François Nezan, Olivier Déforges

Evolutions of RVC so as to handle SVC decoding

14473 Honggang Qi, Wen Gao, Tiejun Huang, Lu Yu

Extension to support non-MPEG standards (ICT/ZJU) (Results of CE 1.6)

14474Honggang Qi, Wen Gao, Lu Yu, Euee S. Jang, Marco Mattavelli, Andrew Kinane

Exploration experiments of AVS decoder description in RVC framework

14480 Paul Schumacher Implementation of MPEG-4 AVC Deblocking Filter in RVC CAL model

14490Christophe Lucarz, Marco Mattavelli, Joseph Thomas-Kerr, Jörn Janneck

Reconfigurability potential of the MPEG-4 SP decoder (results of CE 1.1)

14510 Yuriy Reznik Proposal for adding ISO/IEC 23002-2 in RVC tool library

14546 Jorn JanneckMarco Mattavelli

Description of Tools for the RVC framework: editors, simulator, SW and HDL code generator

Output Documents:

119

No. Title TBP Available23001-4 Codec Configuration Representation

8979 WD 4 of ISO/IEC 23001-4 No 07/05/0423002-4 Video Tool Library

8984 WD 4 of ISO/IEC 23002-4 No 07/05/258985 Description of Core Experiments in RVC No 07/05/048986 RVC Simulation Model (RSM) V4.0 No 07/05/258987 RVC Work Plan No 07/05/048988 RVC Conformance Testing Working Draft 1.0 No 07/05/148989 Description of Exploration Experiments for Toolbox Extensions No 07/05/14

120

Annex I– JVT report

Source: Jens-Rainer Ohm, Gary J. Sullivan, Thomas Wiegand, and Ajay Luthra

1 AbstractThe Joint Video Team (JVT) of ITU-T Q.6/16 and ISO/IEC JTC 1/SC 29/WG 11 held its 23rd meeting during April 21-27, 2007 in San Jose, CA, USA. The JVT meeting was held under the chairmanship of Dr. Gary Sullivan (Microsoft/USA) and Dr. Jens-Rainer Ohm (RWTH Aachen/Germany), and under the associate chairmanship of Dr. Thomas Wiegand (Fraunhofer HHI/Germany) and Dr. Ajay Luthra (Motorola/USA). The JVT meetings opened at approximately 14:30 on Saturday 21 April 2007 and closed at approximately 13:50 on Friday 27 April 2007. Approximately 185 people attended the JVT meetings and approximately 130 input documents were discussed. The meetings took place in a co-located fashion with a meeting of ISO/IEC JTC 1/SC 29/WG 11 (MPEG) – one of the two parent bodies of the JVT. The subject matter of the JVT meeting activities consisted of work on video coding.

49 Contents1. Abstract 1192. Contents 1193. Documents of the JVT meeting 125

3.1. Input documents 1253.1.1 Administrative input contributions..........................................................................1253.1.2 Input liaison statements, WG 11 NB inputs and other noted WG 11 inputs...........1253.1.3 Non-administrative input contributions...................................................................1253.1.4 Late-registered input contributions..........................................................................129

3.2. Late document availability 1293.3. Withdrawn document registrations 1303.4. Major output documents 130

JVT-W200 Meeting report of the 23rd JVT meeting [07/05/20]...............................................................130JVT-W201-M (WG 11 N8962) Joint Draft 10: Scalable Video Coding [07/05/31]..................................130JVT-W202-M (WG 11 N8963) Joint Scalable Video Model (JSVM) 10 [07/05/31]................................130JVT-W203-M (WG 11 N8964) JSVM 10 Software [07/06/29].................................................................130JVT-W204-M (WG 11 N8955) WD 1 conformance test spec for Prof Prof (Teruhiko Suzuki) [07/06/29].....................................................................................................................................................................131JVT-W205-M (WG 11 N8957) WD 1 conformance test for SVC (V. Bottreau) [07/06/29].....................131JVT-W206-M (WG 11 N8959) WD reference software for Prof Prof [07/06/29].....................................131JVT-W207-M (WG 11 N8967) Joint Multi-view Video Model (JMVM) 4 [07/05/18]............................131JVT-W208-M (WG 11 N8968) JMVM 4 Software [07/05/31].................................................................131JVT-V209-M (WG 11 N8966) Joint draft 3 Multi-view Video Coding [07/02/09]...................................131JVT-W211-M (WG 11 N8961) WD reference software for SVC [07/06/29]............................................131JVT-W212-M (WG 11 N8965) Verification test plan for SVC [07/05/18]...............................................131

3.5. JVT internal output documents 131JVT-W210-M ITU-T Rec. H.264 | ISO/IEC 14496-10 Advanced Video Coding Defect Report [07/06/18].....................................................................................................................................................................131

3.6. SVC core experiment output documents 131JVT-W301 CE 1 on SVC subband techiques.............................................................................................131JVT-W302 CE 2 on SVC bit depth and chroma format scalability............................................................131

3.7. MVC core experiment output documents 132JVT-W303 CE 3 on MVC view interpolation/synthesis............................................................................132

4. JVT administrative and liaison topics1324.1. IPR policy reminder and update 132

121

4.2. Meeting opening remarks by the chairmen 1334.3. JVT communication practices1334.4. Scheduling and logistics notes 1344.5. Closing session notes 1344.6. Administrative documents 134

JVT-W001 (Admin) [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] AHG Report: Proj mgmt and errata...........................................................................................................................................................134JVT-W002 (Admin) [K. Suehring, A. Tourapis, T. Suzuki] AHG Report: JM text, ref soft, bitstream, conf.....................................................................................................................................................................137JVT-W003 (Admin) [T. Suzuki] AHG Report: Professional applications.................................................139JVT-W004 (Admin) [J.-R. Ohm, T. Wiegand, M. Bober] AHG Report: Video annotation......................140JVT-W005 (Admin) [G. J. Sullivan, J. Luo] AHG Report: AVC splicing.................................................141JVT-W006 (Admin) [J. Vieron, M. Wien, H. Schwarz, T. Wiegand] AHG Report: JD & JSVM text and S/W.............................................................................................................................................................141JVT-W007 (Admin) [A. Segall, S. Regunathan] AHG Report: SS resampling.........................................144JVT-W008 (Admin) [H. Schwarz, S. Regunathan, A. Eleftheriadis] AHG Report: SVC complexity reduction.....................................................................................................................................................144JVT-W009 (Admin) [Y.-K. Wang, S. Pateux, P. Amon, T. Schierl] AHG Report: SVC high-level syntax, err resil........................................................................................................................................................145JVT-W010 (Admin) [Y. Gao, A. Segall, T. Wiegand] AHG Report: SVC bit depth and chroma format.145JVT-W011 (Admin) [A. Vetro, P. Pandit] AHG Report: MVC high-level syntax & buffering................145JVT-W012 (Admin) [H.-S. Koo] AHG Report: MVC motion/disparity vector coding.............................146JVT-W013 (Admin) [H. Kimata, A. Smolic, P. Pandit, A. Vetro, C. Ying] AHG Report: JMVM & JD text editing.........................................................................................................................................................146JVT-W014 (Admin) [H. Kimata, A. Smolic] AHG Report: MVC exper. framework & test cond............148

4.7. JVT liaison communications 148M14548 WG 11 input [FLO Forum] Liaison statement from FLO Forum to WG 11................................148

5. Scalable video coding 1485.1. CE 1 & related docs: SVC FGS simplification 148

JVT-W090 ( Prop 2.2/3.1) [H. Kirchhoffer, H. Schwarz, T. Wiegand] CE1: Simplified FGS..............148JVT-W115-QV (Late Info) [A. Segall] CE1: Verif JVT-W090 simplified FGS.......................................149JVT-W111 ( Prop 2.2) [M. Karczewicz, S. Park, H. Chung] CE1: Report on FGS simplif...................149JVT-W124-QV (Late Info) [J. Ridge] CE1: Verif JVT-W111 FGS simplif..............................................149JVT-W121 ( Prop 2.2.1/3.1) [J. Ridge, X. Wang] CE1: FGS refinement pass simplif...........................150JVT-W119 ( Prop 2.0/3.1) [Y. Bao, M. Karczewicz, X. Wang, J. Ridge, Y. Ye, W. J. Han, S. Y. Kim] CE1: FGS simplif........................................................................................................................................150JVT-W120 ( Info) [P. Yin] CE1: Verif JVT-W119 FGS simplif............................................................150

5.2. CE 2 & related docs: SVC ESS improvement 150JVT-W030 ( Prop 2.2.1) [X. Wang, J. Ridge] CE2: Improvement of MB mode pred in ESS...............150JVT-W058 ( Info) [E. Francois] CE2: Cross-check of JVT-W030 on ESS mode pred improvement. . .151JVT-W117 ( Prop 2.2/3.1) [Y. Ye, Y. Bao] CE2: Improved resid upsampling for ESS.........................151JVT-W106-QV (Late Info) [X. Wang] CE2: Verif Qualcomm JVT-W117 improved resid upsamp for ESS.....................................................................................................................................................................152JVT-W105 ( Prop 2.0) [X. Wang, J. Ridge] Study on residual upsampling without block boundary check under ESS....................................................................................................................................................152JVT-W109-LV (Late Info) [E. Francois] Verif JVT-W105 on residual upsampling without block boundary check under ESS.........................................................................................................................................153JVT-W123 ( Prop NN2.2.1) [X. Wang, J. Ridge] Analysis of visual artifacts in ESS residual pred.....154

5.3. CE 3 & related docs: SVC subband coding 154JVT-W097 ( Prop 2.2/3.1) [S.-T. Hsiang] CE3: Intra-frame dyadic spatial SVC based on subband/wavelet filter banks framework....................................................................................................154JVT-W122-QV (Late Info) [J. Ridge] CE3: Verif JVT-W097 wavelet-based intra dyadic spatial SVC. .154

5.4. CE 4 & related docs: SVC bit-depth scalability 155JVT-W102 ( Prop 2.2/3.1) [Y. Gao, Y. Wu] CE4: SVC bit-depth scalability simulation results...........155JVT-W116 ( Info) [A. Segall] CE4: Verif JVT-W102 (Thomson prop).................................................156JVT-W113 ( Prop 2.2) [A. Segall, Y. Su] System for bit-depth scalable coding....................................156JVT-W076 ( Prop 2.2) [J. Jia, H. K. Kim, H. C. Choi, J. J. Yoo] SVC chroma format scalability........156

5.5. SVC high-level syntax 156JVT-W020 ( Prop 2.0) [Z. G. Li, S. Rahardja, S. L. Xie and W. Yao] Hypothetical reference decoder for video coding................................................................................................................................................156

122

JVT-W046 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Support for SVC header rewriting to AVC............................................................................................................................................................158JVT-W047 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Pictures not for output in SVC...........158JVT-W048 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang, Y. Chen] On SVC high-level syntax..159JVT-W051 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] On SVC scalability information related SEI messages...................................................................................................................................159JVT-W052 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] SVC feedback based coding..............159JVT-W137-B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Revised syntax for quality layer SEI message................................................................................................................................................160JVT-W053 ( Prop 2.0/3.1) [M. M. Hannuksela, Y.-K. Wang, D. Singer, T. Rathgen] SVC priority_id value setting method indication..................................................................................................................160JVT-W064 ( Prop 2.2/3.1) [J. Luo, L. Zhu, P. Yin, C. Gomila] VUI updates for SVC..........................160JVT-W091 ( Prop 2.2/3.1) [L. Cieplinski] HRD parameters for SVC bitstream rewriting....................161JVT-W114 ( Prop 2.2) [A. Segall, J. Zhao] Showcase for transcoding scalability info SEI..................161JVT-W125 ( Prop 2.2) [G. J. Sullivan] On SVC high-level syntax and HRD........................................161JVT-W049 ( Prop 2.2.1/3.1) [C. He, H. Liu, H. Li, Y.-K. Wang, M. M. Hannuksela] Redundant pictures in SVC.........................................................................................................................................................162JVT-W050 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] On tl0_pic_idx in SVC.......................162JVT-W062 ( Prop 2.2/3.1) [A. Eleftheriadis, S. Cipolli, J. Lennox] Improved error resilience using temporal level 0 picture index.....................................................................................................................163JVT-W054 ( Info) [I. Radulovic, Y.-K. Wang, S. Wenger, A. Hallapuro, M. M. Hannuksela] Multiple description coding using AVC redundant pictures.....................................................................................163JVT-W068 ( Prop 2.2/3.1) [C. Tu, S. Srinivasan, S. Regunathan, G. Sullivan] CE4: 4-tap MC interp for high-res SVC enh layers.............................................................................................................................164JVT-W072 ( Info) [H. Schwarz] Results comparing JSVM, 4-tap, and RCDO MC interp....................165JVT-W027 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of 4 tap motion compensation interp.....................................................................................................................................................................165Discussion of potential rearrangement of NAL unit order.........................................................................165

5.6. SVC applications and profiles 166JVT-W075 ( Prop 2.0/3.1) [M. Horowitz, A. Eleftheriadis] Max frame size for enh layers of SVC profiles <withdrawn>..................................................................................................................................166JVT-W093 ( Prop 2.2.1/3.1) [H. Chung, M. Karczewicz, J. Ridge, X. Wang, W. Han, S. Kim] SVC FGS profile..........................................................................................................................................................166Profiles definition changes..........................................................................................................................167

5.7. SVC other normative design proposals 1695.7.1 SVC restrictions on interlaced coding.....................................................................169

JVT-W025 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Restrictions on interlaced coding in SVC.....................................................................................................................................................................169

5.7.2 SVC smoothed reference prediction........................................................................169JVT-W026 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Profile SVC B: Evaluation of smoothed ref pred........................................................................................................................................................169JVT-W118 ( Prop 2.2) [Y. Ye, Y. Bao, W. J. Han, S. Y. Kim] Perf and complexity of smoothed ref pred.....................................................................................................................................................................170JVT-W126 ( Info) [Z. He] Verif JVT-W118 perf and complexity of smoothed ref pred.......................172JVT-W112-L (Late Prop 2.2) [A. Segall] Clarification of base_mode_flag <withdrawn>.......................172

5.7.3 SVC deblocking.......................................................................................................172JVT-W061 ( Prop 2.2/3.1) [D. Hong, A. Eleftheriadis, O. Shapiro] Modified deblocking filter process in scalable extension.......................................................................................................................................172JVT-W063 ( Prop 2.0/3.1 Layered Media, then 2.2 from Polycom) [D. Hong. A. Eleftheriadis, O. Shapiro] Deblocking filter for SVC to support multi-threading with slice boundary................................174JVT-W069 ( Prop 2.2/3.1) [Z. He] Simplified H.264/AVC deblocking filter for SVC enh layer..........175JVT-W128-QV (Late Info) [Y. Ye] Verif of JVT-W069: Simplified deblocking for SVC enh layer.......175

5.7.4 SVC spatial scalability resampling..........................................................................175JVT-W028 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of flexible 4-tap upsampling filters.....................................................................................................................................................................175JVT-W022 ( Prop 2.2/3.1) [T. Tran, L. Liu, P. Topiwala] Dyadic spatial down- and up-sampling filters for SVC.......................................................................................................................................................176JVT-W086 ( Prop 2.2/3.1) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Some consideration on the up-sampling position calculation................................................................................................................177JVT-W136-B (BoG) [G. J. Sullivan, S. Pateux] BoG report on JVT-W086..............................................177

5.8. SVC non-normative contributions 1775.8.1 SVC editorial input..................................................................................................177

123

JVT-W070 ( Text) [H. Schwarz, M. Wien] Editors input for SVC draft................................................177JVT-W099 ( Info) [J. H. Park, Y. H. Kim, B. H. Choi] Clarification of mb_qp_delta syntax...............177

5.8.2 SVC tutorial material...............................................................................................178JVT-W132-B (Requested Info) [T. Wiegand] Overview paper and presentation on SVC........................178

5.8.3 SVC encoder and extractor optimization.................................................................178JVT-W071 ( Info) [H. Schwarz, T. Wiegand] Further results for an rd-opt. multi-loop SVC enc.........178JVT-W029 ( Info 2.2.1/3.1) [W.-H. Peng] Low-complexity mode decision algorithm for combined CGS and temporal scalability..............................................................................................................................178JVT-W043 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control for the Joint Scalable Video Model (JSVM)........................................................................................................................................................179

5.9. SVC conformance 179JVT-W138-B (BoG) [V. Bottreau] Toward an SVC conformance specification.......................................179

5.10. SVC verification testing 180JVT-W110 ( Info) [E. Francois, V. Bottreau, J. Vieron] SVC verif test plan: Updated results for SVC High Profile intra........................................................................................................................................180JVT-W131-B (Late Info) [D. Hong, A. Eleftheriadis] Verification bitstreams for SVC Profile A...........180JVT-W135-B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] On SVC verif test plan.............180

6. Multi-view coding 1806.1. CE 5 & related docs: MVC illumination compensation 180

JVT-W024 ( Prop 2.2/3.1) [W. S. Shim, M. W. Park, G. H. Park, D. Y. Suh, H. S. Song, Y. H. Moon, J. B. Choi] CE5 results- joint prop for MVC deblocking...............................................................................180JVT-W023 ( Info) [S.-C. Lim, D.-H. Han, Y.-L. Lee] CE5: Verification of loop filtering in MVC......180JVT-W031 ( Prop 2.2) [J.-H. Yang] CE5: Illumination comp. info. derivation for MVC......................181JVT-W085 ( Info) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] CE5: Verification of JVT-W031 illumination comp. info. derivation............................................................................................................181

6.2. CE 6 & related docs: MVC view interpolation 181JVT-W055 ( Prop 2.2/3.1) [T. Senoh, M. Okui, K. Enami] Experimental results of camera-rotation-compensated prediction in CE6..................................................................................................................181JVT-W059 ( Prop 2.2/3.1) [S. Yea, A. Vetro] CE6: View synthesis prediction.....................................182JVT-W084 ( Info 2.2/3.1) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] Observations of multi-view test sequences.....................................................................................................................................182JVT-W083 ( Prop 2.2/3.1) [Y. S. Ho, C. Lee, K. J. Oh, B. H. Choi, J. H. Park] CE6: View interp pred for MVC...........................................................................................................................................................182JVT-W103 ( Info) [J.-H. Yang, S.-H. Lee] CE6: Verif GIST MVC contribution JVT-W083 MVC view interp pred...................................................................................................................................................183JVT-W096 ( Prop 2.2/3.1) [S. Naito, A. Koike] CE6: Results on MVC................................................183JVT-W087 ( Prop 2.2/3.1) [S. Shimizu, H. Kimata] New view synthesis pred framework using resid pred.....................................................................................................................................................................183Anthony Vetro presents new CE6 work plan.............................................................................................184JVT-W133-B (BoG) [A. Vetro] BoG report on MVC view interpolation pred.........................................184

6.3. MVC high-level syntax 184JVT-W035 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] On MVC JD 2.0..................184JVT-W036 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela, Y. Chen] MVC output related conformance................................................................................................................................................185JVT-W037 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] View scalable SEI message for MVC...........................................................................................................................................................185JVT-W038 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] Operation point and view dependency changes SEI messages for MVC.............................................................................................185JVT-W039 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] Non-required pictures SEI message for MVC.......................................................................................................................................186JVT-W056 ( Prop 2.2) [J. B. Choi, W. S. Shim, H. S. Song, Y. H. Moon] Inter-view prediction reference picture marking...........................................................................................................................................186JVT-W066 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Ref pic list reordering for MVC..................186JVT-W067 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] H.264/AVC extension for MVC using SEI message.......................................................................................................................................................187JVT-W074 ( Prop 2.2) [H. S. Song, W. S. Shim, Y. H. Moon, J. B. Choi] Comments on view dependency info..........................................................................................................................................187JVT-W080 ( Prop 2.2.1) [K. Ugur, H. Liu, Y.-K. Wang] Showcase for parallel decoding info SEI message for MVC.......................................................................................................................................187JVT-W088 ( Prop 2.2) [S. Lin, S. Gao, Y. Liu, L. Xiong] H.264/AVC SEI extensions for MVC.........188

6.4. MVC other normative technical inputs 1886.4.1 MVC motion/disparity vector coding......................................................................188

124

JVT-W081 ( Prop 2.2) [H. S. Koo, Y. J. Jeon, B. M. Jeon] MVC motion skip mode............................188JVT-W139-B (BoG) [LG, Thomson] Break-out conclusions on JVT-W081............................................189JVT-W073 ( Info) [K. Sohn, J. Seo] Verification of JVT-W081 LGE MVC motion skip contrib.........189JVT-W101 ( Prop 2.2) [H. Yan, J. Huo, Y. Chang, S. Lin, S. Gao, L. Xiong] MV/DV prediction based on RDV.......................................................................................................................................................189JVT-W104 ( Prop 2.2) [S.-H. Lee, S.-H. Lee, N.-I. Cho, J.-H. Yang] MVC disparity vector pred........189JVT-W107 ( Info) [K. Sohn, J. Seo] Verif JVT-W104 MVC disparity vector pred...............................189

6.4.2 MVC weighted prediction.......................................................................................190JVT-W040 ( Prop 2.2.1/3.1) [S. Liu, Y. Chen, Y.-K. Wang, M. M. Hannuksela] Constraints on temporal direct mode and weighted prediction in MVC............................................................................................190JVT-W098 ( Prop 2.2) [J. H. Park, Y. H. Kim, J. W. Kim, B. H. Choi] Weighted prediction for MVC190

6.4.3 MVC downsampled reference etc...........................................................................190JVT-W079 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Inter-view prediction with downsampled reference pictures.........................................................................................................190JVT-W092 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Reduced resolution update for MVC..........191JVT-W094 ( Info 2.2) [W. J. Tam] Image and depth quality of asymmetrically coded stereoscopic video for 3D-TV...................................................................................................................................................191

6.4.4 MVC modes and other coding efficiency topics.....................................................192JVT-W078 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Co-located block condition for inter-view prediction.............................................................................................................192JVT-W082 ( Prop 2.2) [Y. J. Jeon, H. S. Koo, B. M. Jeon] Modified spatial direct mode in MVC......192JVT-W065 ( Prop 2.2/3.1) [P. L. Lai, A. Ortega, P. Pandit, P. Yin, C. Gomila] Adaptive reference filtering for MVC........................................................................................................................................192

6.4.5 MVC depth-based methods & displays...................................................................193JVT-W095 ( Info 2.2) [W. J. Tam, L. Zhang] Depth map preproc and minimal content for 3D-TV using depth-based rendering.................................................................................................................................193JVT-W100 ( Prop 2.0/3.1) [A. Smolic, K. Mueller, P. Merkle, N. Atzpadin, C. Fehn, M. Mueller, O. Schreer, R. Tanger, P. Kauff, T. Wiegand, T. Balogh, Z. Megyesi, A. Barsi] Multi-view video plus depth (MVD) format for advanced 3D video systems..........................................................................................193JVT-W060 ( Prop 2.2/3.1) [A. Vetro, S. Yea, W. Matusik, H. Pfister, M. Zwicker] Anti-aliasing for 3D displays.......................................................................................................................................................195

6.4.6 MVC view parallel processing................................................................................195JVT-W077 ( Prop 2.2) [P. Yang, X. Xu, G. Zhu, Y. He] View parallel processing on MVC................195JVT-W108-QV (Late Info) [Q. Chen, Z. Chen] Verif JVT-W077 view parallel proc on MVC................196

6.5. MVC reference software, common conditions, encoder optimization 1967. AVC base specification and related topics 196

JVT-W041 ( Prop NN) [A. M. Tourapis, K. Suehring, G. J. Sullivan, A. Leontaris] H.264/MPEG-4 AVC reference software (JM) manual.................................................................................................................196JVT-W042 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control reorganization in the JM reference software.......................................................................................................................................................197JVT-W044 ( Info) [A. M. Tourapis, A. Leontaris, K. Suehring] New JM reference software enhancements..............................................................................................................................................197JVT-W057 (Late Info) [K. P. Lim] Improved JM text algorithm description............................................198JVT-W140-B (BoG) [T. Suzuki] Toward a professional profiles conformance specification...................198

8. Video annotation (jointly discussed with MPEG requirements 3:30 pm Wednesday 25 April) 198

JVT-W032 ( Info) [Q. Chen, C. Louis, Z. Chen] Requirements of video annotation in video coding...198JVT-W033 ( Prop 2.2/3.1) [Q. Chen, Z. Chen, X. Gu] Video annotation SEI message.........................199JVT-W034 ( Prop 2.2/3.1) [C. Louis, O. Lionel, L. Frederic, Z. Chen, Q. Chen] Fingerprint and video structure for video annotation SEI message................................................................................................199

9. AVC errata and clarification issues 200JVT-W134-Q (Late Prop 2.2) [S. Narasimhan] Splicing issues and some suggested changes..................200

10. Requirements joint discussions with WG 11 200M14452 WG 11 input [T. Murakami, K. Asai, Y. Yamada] Requirement of full-color video coding for consumer applications.................................................................................................................................200M14360 [USNB to WG 11] Issues relating to expiring patents.................................................................201JVT-W127 ( Req) [M. Tanimoto, T. Fujii, H. Kimata, S. Sakazawa] Requirements for FTV (MPEG M14417)......................................................................................................................................................201

11. JVT internal operating rules 20212. List of adoptions 204

12.1. SVC normative adoptions into JD 204125

12.2. SVC normative adoptions into JSVM 20412.3. SVC non-normative adoptions 20412.4. SVC software adoptions 20512.5. MVC normative JD adoptions 20512.6. MVC JMVM adoptions 20512.7. MVC non-normative adoptions 20512.8. JM non-normative adoptions 20512.9. Other normative adoptions 20512.10. Other non-normative adoptions 205

13. Software integration plan 20514. SVC conformance work plan 20515. SVC verification test plan 20616. List of AHGs established 206

16.1. JVT project management and errata reporting 20616.2. JM Text, reference software, bitstream exchange and conformance 20616.3. AVC professional applications 20616.4. SVC JD and JSVM text, software and conformance 20716.5. SVC bit depth and chroma format scalability20716.6. SVC FGS applications and design simplification 20716.7. MVC high-level syntax and buffer management 20716.8. MVC JD and JSVM text and software 20716.9. MVC experimental framework and testing conditions 20816.10. MVC solutions using existing AVC decoders 20816.11. MVC reduced resolution update, downsampled reference and adaptive reference filtering 208

17. Resolutions conveyed to MPEG parent body 20817.1. Resolutions relating to ISO/IEC 14496-4 208

17.1.1 The JVT and the video subgroup recommend to approve the following documents208

17.1.2 The JVT and the video subgroup thank the following companies for their commitment to provide conformance testing streams for ISO/IEC 14496-4:2004/Amd.30: Mitsubishi Electric Corp., Panasonic, Sejong University, Thomson..................................20917.1.3 The JVT and the video subgroup thank the following companies for their commitment to provide conformance testing streams for ISO/IEC 14496-4:2004/Amd.31: ETRI, FhG-HHI, France Telecom/Orange, Layered Media, Sharp, Thomson...................209

17.2. Resolutions relating to ISO/IEC 14496-5 20917.2.1 The JVT and the video subgroup recommend to approve the following documents

20917.3. Resolutions relating to ISO/IEC 14496-10 209

17.3.1 The JVT and the video subgroup recommend to approve the following documents209

17.3.2 The JVT and the video subgroup request WG 11 National Bodies to kindly consider the SVC Study Document N8962 [JVT-W201] and if necessary provide additional comments by the July 2007 meeting...................................................................................209

17.4. Resolutions relating to future meeting scheduling 20917.4.1 The JVT chairmen propose to hold a JVT meeting during June 29 through July 6, 2007 under the auspices of the meeting of ITU-T SG 16 in Geneva, CH. Further meetings are proposed to be held during October 19-26, 2007 under WG 11 auspices in Shenzhen, CN, and during January 11-18, 2008 under WG 11 auspices in Antalya, TR.....................209

17.5. Resolutions relating to ad hoc group activities 21017.5.1 The JVT provides the following list of JVT ad hoc groups appointed to progress work in the interim period until the next JVT meeting:......................................................210

18. Attendance 210

126

50 Documents of the JVT meeting

50.1 Input documents

50.1.1 Administrative input contributionsJVT-W000 (Admin) List of documents of San Jose meetingJVT-W001 (Admin) [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] AHG Report: Proj mgmt

and errataJVT-W002 (Admin) [K. Suehring, A. Tourapis, T. Suzuki] AHG Report: JM text, ref soft,

bitstream, confJVT-W003 (Admin) [T. Suzuki] AHG Report: Professional applicationsJVT-W004 (Admin) [J.-R. Ohm, T. Wiegand, M. Bober] AHG Report: Video annotationJVT-W005 (Admin) [G. J. Sullivan, J. Luo] AHG Report: AVC splicingJVT-W006 (Admin) [J. Vieron, M. Wien, H. Schwarz, T. Wiegand] AHG Report: JD & JSVM

text and S/WJVT-W007 (Admin) [A. Segall, S. Regunathan] AHG Report: SS resamplingJVT-W008 (Admin) [H. Schwarz, S. Regunathan, A. Eleftheriadis] AHG Report: SVC

complexity reductionJVT-W009 (Admin) [Y.-K. Wang, S. Pateux, P. Amon, T. Schierl] AHG Report: SVC high-level

syntax, err resilJVT-W010 (Admin) [Y. Gao, A. Segall, T. Wiegand] AHG Report: SVC bit depth and chroma

formatJVT-W011 (Admin) [A. Vetro, P. Pandit] AHG Report: MVC high-level syntax & bufferingJVT-W012 (Admin) [H.-S. Koo] AHG Report: MVC motion/disparity vector codingJVT-W013 (Admin) [H. Kimata, A. Smolic, P. Pandit, A. Vetro, C. Ying] AHG Report: JMVM

& JD text editingJVT-W014 (Admin) [H. Kimata, A. Smolic] AHG Report: MVC exper. framework & test cond

50.1.2 Input liaison statements, WG 11 NB inputs and other noted WG 11 inputsThe following input documents to WG 11 were noted by the JVT and discussed jointly with WG 11 (without JVT action).

M14360 WG 11 input [USNB to WG 11] Issues relating to expiring patentsM14452 WG 11 input [T. Murakami, K. Asai, Y. Yamada] Requirement of full-color video

coding for consumer applicationsM14548 WG 11 input [FLO Forum] Liaison statement from the FLO Forum

50.1.3 Non-administrative input contributionsJVT-W020 ( Prop 2.0) [Z. G. Li, S. Rahardja, S. L. Xie and W. Yao] Hypothetical reference

decoder for video codingJVT-W021 [withdrawn] <withdrawn>JVT-W022 ( Prop 2.2/3.1) [T. Tran, L. Liu, P. Topiwala] Dyadic spatial down- and up-

sampling filters for SVCJVT-W023 ( Info) [S.-C. Lim, D.-H. Han, Y.-L. Lee] CE5: Verification of loop filtering in

MVCJVT-W024 ( Prop 2.2/3.1) [W. S. Shim, M. W. Park, G. H. Park, D. Y. Suh, H. S. Song, Y. H.

Moon, J. B. Choi] CE5 results- joint prop for MVC deblockingJVT-W025 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Restrictions on interlaced

coding in SVC

127

JVT-W026 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Profile SVC B: Evaluation of smoothed ref pred

JVT-W027 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of 4 tap motion compensation interp

JVT-W028 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of flexible 4-tap upsampling filters

JVT-W029 ( Info 2.2.1/3.1) [W.-H. Peng] Low-complexity mode decision algorithm for combined CGS and temporal scalability

JVT-W030 ( Prop 2.2.1) [X. Wang, J. Ridge] CE2 report: Improvement of macroblock mode prediction in ESS

JVT-W031 ( Prop 2.2) [J.-H. Yang] CE5: Illumination comp. info. derivation for MVCJVT-W032 ( Info) [Q. Chen, C. Louis, Z. Chen] Requirements of video annotation in video

codingJVT-W033 ( Prop 2.2/3.1) [Q. Chen, Z. Chen, X. Gu] Video annotation SEI messageJVT-W034 ( Prop 2.2/3.1) [C. Louis, O. Lionel, L. Frederic, Z. Chen, Q. Chen] Fingerprint and

video structure for video annotation SEI messageJVT-W035 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] Comments to MVC

JD 2.0JVT-W036 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela, Y. Chen] MVC output related

conformanceJVT-W037 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] View scalable SEI

message for MVCJVT-W038 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] Operation point and

view dependency changes SEI messages for MVCJVT-W039 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] Non-required pictures

SEI message for MVCJVT-W040 ( Prop 2.2.1/3.1) [S. Liu, Y. Chen, Y.-K. Wang, M. M. Hannuksela] Constraints on

temporal direct mode and weighted prediction in MVCJVT-W041 ( Prop NN) [A. M. Tourapis, K. Suehring, G. J. Sullivan, A. Leontaris]

H.264/MPEG-4 AVC reference software (JM) manualJVT-W042 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control reorganization in the JM

reference softwareJVT-W043 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control for the Joint Scalable

Video Model (JSVM)JVT-W044 ( Info) [A. M. Tourapis, A. Leontaris, K. Suehring] New JM reference software

enhancementsJVT-W045 [withdrawn] <withdrawn>JVT-W046 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Support for SVC header

rewriting to AVCJVT-W047 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Pictures not for output in SVCJVT-W048 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang, Y. Chen] On SVC high-level

syntaxJVT-W049 ( Prop 2.2.1/3.1) [C. He, H. Liu, H. Li, Y.-K. Wang, M. M. Hannuksela] Redundant

pictures in SVCJVT-W050 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] On tl0_pic_idx in SVCJVT-W051 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] On SVC scalability

information related SEI messagesJVT-W052 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] SVC feedback based codingJVT-W053 ( Prop 2.0/3.1) [M. M. Hannuksela, Y.-K. Wang, D. Singer, T. Rathgen] SVC

priority_id value setting method indicationJVT-W054 ( Info) [I. Radulovic, Y.-K. Wang, S. Wenger, A. Hallapuro, M. M. Hannuksela]

Multiple description coding using AVC redundant pictures

128

JVT-W055 ( Prop 2.2/3.1) [T. Senoh, M. Okui, K. Enami] Experimental results of camera-rotation-compensated prediction in CE6

JVT-W056 ( Prop 2.2) [J. B. Choi, W. S. Shim, H. S. Song, Y. H. Moon] Inter-view prediction reference picture marking

JVT-W057 (Late Info) [K. P. Lim] Improved JM text algorithm descriptionJVT-W058 ( Info) [E. Francois] CE2: Cross-check of JVT-W030 on ESS mode pred

improvementJVT-W059 ( Prop 2.2/3.1) [S. Yea, A. Vetro] CE6: View synthesis predictionJVT-W060 ( Prop 2.2/3.1) [A. Vetro, S. Yea, W. Matusik, H. Pfister, M. Zwicker] Anti-

aliasing for 3D displaysJVT-W061 ( Prop 2.2/3.1) [D. Hong, A. Eleftheriadis, O. Shapiro] Modified deblocking filter

process in scalable extensionJVT-W062 ( Prop 2.2/3.1) [A. Eleftheriadis, S. Cipolli, J. Lennox] Improved error resilience

using temporal level 0 picture indexJVT-W063 ( Prop 2.0/3.1, then 2.2) [D. Hong. A. Eleftheriadis, O. Shapiro] Deblocking filter

for SVC to support multi-threading with slice boundaryJVT-W064 ( Prop 2.2/3.1) [J. Luo, L. Zhu, P. Yin, C. Gomila] VUI updates for SVCJVT-W065 ( Prop 2.2/3.1) [P. L. Lai, A. Ortega, P. Pandit, P. Yin, C. Gomila] Adaptive

reference filtering for MVCJVT-W066 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Ref pic list reordering for MVCJVT-W067 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] H.264/AVC extension for MVC

using SEI messageJVT-W068 ( Prop 2.2/3.1) [C. Tu, S. Srinivasan, S. Regunathan, G. J. Sullivan] CE4: 4-tap MC

interp for high-res SVC enh layersJVT-W069 ( Prop 2.2/3.1) [Z. He] Simplified H.264/AVC deblocking filter for SVC enh layerJVT-W070 ( Text) [H. Schwarz, M. Wien] Editors input for SVC draftJVT-W071 ( Info) [H. Schwarz, T. Wiegand] Further results for an rd-opt. multi-loop SVC enc.JVT-W072 ( Info) [H. Schwarz] Results comparing JSVM, 4-tap, and RCDO MC interp.JVT-W073 ( Info) [K. Sohn, J. Seo] Verification of JVT-W081 LGE MVC motion skip

contrib.JVT-W074 ( Prop 2.2) [H. S. Song, W. S. Shim, Y. H. Moon, J. B. Choi] Comments on view

dependency infoJVT-W075 ( Prop 2.0/3.1) [M. Horowitz, A. Eleftheriadis] Max frame size for enh layers of

SVC profiles <withdrawn>JVT-W076 ( Prop 2.2) [J. Jia, H. K. Kim, H. C. Choi, J. J. Yoo] SVC chroma format scalabilityJVT-W077 ( Prop 2.2) [P. Yang, X. Xu, G. Zhu, Y. He] View parallel processing on MVCJVT-W078 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Co-located

block condition for inter-view predictionJVT-W079 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Inter-view

prediction with downsampled reference picturesJVT-W080 ( Prop 2.2.1) [K. Ugur, H. Liu, Y.-K. Wang] Showcase for parallel decoding

information SEI message for MVCJVT-W081 ( Prop 2.2) [H. S. Koo, Y. J. Jeon, B. M. Jeon] MVC motion skip modeJVT-W082 ( Prop 2.2) [Y. J. Jeon, H. S. Koo, B. M. Jeon] Modified spatial direct mode in

MVCJVT-W083 ( Prop 2.2/3.1) [Y. S. Ho, C. Lee, K. J. Oh, B. H. Choi, J. H. Park] CE6: View

interp pred for MVCJVT-W084 ( Info 2.2/3.1) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] Observations of

multi-view test sequencesJVT-W085 ( Info) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] CE5: Verification of

JVT-W031 illumination comp. info. derivationJVT-W086 ( Prop 2.2/3.1) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Some

consideration on the up-sampling position calculation

129

JVT-W087 ( Prop 2.2/3.1) [S. Shimizu, H. Kimata] New view synthesis prediction framework using residual prediction

JVT-W088 ( Prop 2.2) [S. Lin, S. Gao, Y. Liu, L. Xiong] H264/AVC SEI extensions for MVCJVT-W089 [withdrawn] <withdrawn>JVT-W090 ( Prop 2.2/3.1) [H. Kirchhoffer, H. Schwarz, T. Wiegand] CE1: Simplified FGSJVT-W091 ( Prop 2.2/3.1) [L. Cieplinski] HRD parameters for SVC bitstream rewritingJVT-W092 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Reduced resolution update for MVCJVT-W093 ( Prop 2.2.1/3.1) [H. Chung, M. Karczewicz, J. Ridge, X. Wang, W. Han, S. Kim]

SVC FGS profileJVT-W094 ( Info 2.2) [W. J. Tam] Image and depth quality of asymmetrically coded

stereoscopic video for 3D-TVJVT-W095 ( Info 2.2) [W. J. Tam, L. Zhang] Depth map preproc and minimal content for 3D-

TV using depth-based renderingJVT-W096 ( Prop 2.2/3.1) [S. Naito, A. Koike] CE6: Results on MVCJVT-W097 ( Prop 2.2/3.1) [S.-T. Hsiang] CE3: Intra-frame dyadic spatial SVC based on

subband/wavelet filter banks frameworkJVT-W098 ( Prop 2.2) [J. H. Park, Y. H. Kim, J. W. Kim, B. H. Choi] Weighted prediction for

MVCJVT-W099 ( Info) [J. H. Park, Y. H. Kim, B. H. Choi] Clarification of mb_qp_delta syntaxJVT-W100 ( Prop 2.0/3.1) [A. Smolic, K. Mueller, P. Merkle, N. Atzpadin, C. Fehn, M.

Mueller, O. Schreer, R. Tanger, P. Kauff, T. Wiegand, T. Balogh, Z. Megyesi, A. Barsi] Multi-view video plus depth (MVD) format for advanced 3D video systems

JVT-W101 ( Prop 2.2) [H. Yan, J. Huo, Y. Chang, S. Lin, S. Gao, L. Xiong] MV/DV prediction based on RDV

JVT-W102 ( Prop 2.2/3.1) [Y. Gao, Y. Wu] CE4: SVC bit-depth scalability simulation resultsJVT-W103 ( Info) [J.-H. Yang, S.-H. Lee] CE6: Verif GIST MVC contribution JVT-W083

MVC view interp predJVT-W104 ( Prop 2.2) [S.-H. Lee, S.-H. Lee, N.-I. Cho, J.-H. Yang] MVC disparity vector

predJVT-W105 ( Prop 2.0) [X. Wang, J. Ridge] Study on residual upsampling without block

boundary check under ESSJVT-W106-QV (Late Info) [X. Wang] CE2: Verif Qualcomm JVT-W117 improved resid upsamp

for ESSJVT-W107 ( Info) [K. Sohn, J. Seo] Verif JVT-W104 MVC disparity vector predJVT-W108-Q (Late Info) [Q. Chen, Z. Chen] Verif JVT-W077 view parallel proc on MVCJVT-W109-LV (Late Info) [E. Francois] Cross-check of JVT-W105 on residual upsampling

without block boundary check under ESSJVT-W110 ( Info) [E. Francois, V. Bottreau, J. Vieron] SVC verif test plan: Updated results for

SVC High Profile intraJVT-W111 ( Prop 2.2) [M. Karczewicz, S. Park, H. Chung] CE1: Report on FGS simplifJVT-W112-L (Late Prop 2.2) [A. Segall] Clarification of base_mode_flag <withdrawn>JVT-W113 ( Prop 2.2) [A. Segall, Y. Su] System for bit-depth scalable codingJVT-W114 ( Prop 2.2) [A. Segall, J. Zhao] Showcase for transcoding scalability info SEIJVT-W115-QV (Late Info) [A. Segall] CE1: Verif JVT-W090 simplified FGSJVT-W116 ( Info) [A. Segall] CE4: Verif JVT-W102 (Thomson prop)JVT-W117 ( Prop 2.2/3.1) [Y. Ye, Y. Bao] CE2: Improved resid upsampling for ESSJVT-W118 ( Prop 2.2) [Y. Ye, Y. Bao, W. J. Han, S. Y. Kim] Perf and complexity of smoothed

ref predJVT-W119 ( Prop 2.0/3.1) [Y. Bao, M. Karczewicz, X. Wang, J. Ridge, Y. Ye, W. J. Han, S.

Y. Kim] CE1: FGS simplifJVT-W120 ( Info) [P. Yin] CE1: Verif JVT-W119 FGS simplifJVT-W121 ( Prop 2.2.1/3.1) [J. Ridge, X. Wang] CE1: FGS refinement pass simplif

130

JVT-W122-QV (Late Info) [J. Ridge] CE3: Verif JVT-W097 wavelet-based intra dyadic spatial SVC

JVT-W123 ( Prop NN2.2.1) [X. Wang, J. Ridge] Analysis of visual artifacts in ESS residual pred

JVT-W124-QV (Late Info) [J. Ridge] CE1: Verif JVT-W111 FGS simplifJVT-W125 ( Prop 2.2) [G. J. Sullivan] On SVC high-level syntax and HRDJVT-W126 ( Info) [Z. He] Verif JVT-W118 perf and complexity of smoothed ref predJVT-W127 ( Req) [M. Tanimoto, T. Fujii, H. Kimata, S. Sakazawa] Requirements for FTV

(MPEG M14417)

50.1.4 Late-registered input contributionsJVT-W128-QV (Late Info) [Y. Ye] Verif of JVT-W069: Simplified deblocking for SVC enh

layerJVT-W129 [withdrawn] <withdrawn>JVT-W130 [withdrawn] <withdrawn>JVT-W131-B (Late Info) [D. Hong, A. Eleftheriadis] Verification bitstreams for SVC Profile AJVT-W132-B (Requested Info) [T. Wiegand] Overview paper and presentation on SVCJVT-W133-B (BoG) [A. Vetro] BoG report on MVC view interpolation predJVT-W134-Q (Late Prop 2.2) [S. Narasimhan] Splicing issues and some suggested changesJVT-W135-B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] On SVC verif test planJVT-W136-B (BoG) [G. J. Sullivan, S. Pateux] BoG report on JVT-W086JVT-W137-B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Revised syntax for

quality layer SEI messageJVT-W138-B (BoG) [V. Bottreau] Toward an SVC conformance specificationJVT-W139-B (BoG) [LG, Thomson] Break-out conclusions on JVT-W081JVT-W140-B (BoG) [T. Suzuki] Toward a professional profiles conformance specification

50.2 Late document availabilityDocument numbers suffixed in this report with "-L", "-Q", or "-M" were classified as late. Such documents will only be considered as information documents only (unless agreed otherwise by the group) if time permits, and consideration of them may be shifted to the end of the meeting as determined appropriate by the group.

Furthermore, due to our difficulties with a large quantity of late-submitted contributions at recent previous meetings, the JVT agreed at its preceding meeting that for this meeting, no late-uploaded (non-AHG-report, non-liaison) contribution would be presented without having a minimum of 4 JVT participants (working for organizations other than that of the primary contribution author) recorded by name as supporting the allowance of such a presentation, in addition to a consensus of the general JVT membership to allow the presentation. Such support to allow a presentation is to be understood to not necessarily imply support of the adoption of the content of the late contribution, but only as a positive expression that the document should be allowed to be presented. Additionally, the provider of a presented late contribution shall send an email apology to the JVT email reflector. This rule does not apply to material requested by the JVT at the meeting (e.g., reports of JVT-authorized side activities).

Clarification: Does not apply to verification contributions.

Further clarification: The four people shall be from different organizations.

JVT decision: Agreed.

131

A check mark () indicates a contribution considered to be available on time.

The suffixes for contributions not marked as “” are explained below:– "-L" indicates a contribution that was somewhat late but was available by the first meeting

day.– "-Q" were more late than that.– "-M" were still missing at the time of preparation of this report.– "-B" were break-out group discussion reports and other input requested during the meeting

Further suffixing by “V” indicates a verification contribution.

Contribution JVT-W134 (from S. Narasimhan) was subject to lateness penalties. An apology for the lateness of the contribution was sent to the JVT email reflector, and JVT members were recorded by name requesting presentation as follows: Mukta Kar, Jian Zong, Katie Cornog, and Wade Wan. Presentation of JVT-W134 was postponed to the last meeting day, and no immediate action was taken in response to the contribution (other than to include it in a list of issues to be considered for later action).

There were no objections to presentations of late documents at this meeting.

JVT-W112 (from A. Segall) was also late. Although supported for presentation by four JVT members (requesting presentation: Miska Hannuksela, Mathias Wien, Peter Amon, Vincent Bottreau), the contribution was withdrawn as moot after some discussion, in consideration of action taken in response to other contributions.

It was noted that, with only one (non-withdrawn) contribution subject to lateness penalties (and that one having no immediate action requested or taken), the situation surrounding the need for on-time availability of contributions has substantially improved.

50.3 Withdrawn document registrationsJVT-W021 [withdrawn] <withdrawn>JVT-W045 [withdrawn] <withdrawn>JVT-W075 [M. Horowitz, A. Eleftheriadis] Max frame size for enh layers of SVC profiles <withdrawn>JVT-W089 [withdrawn] <withdrawn>JVT-W112-L [A. Segall] Clarification of base_mode_flag <withdrawn>JVT-W129 [withdrawn] <withdrawn>JVT-W130 [withdrawn] <withdrawn>

50.4 Major output documentsMajor output documents submitted to parent-body review included the following. (Dates listed are planned dates of availability.)

50.4.1.1.1 JVT-W200 Meeting report of the 23rd JVT meeting [07/05/20]

50.4.1.1.2 JVT-W201-M (WG 11 N8962) Joint Draft 10: Scalable Video Coding [07/05/31]

50.4.1.1.3 JVT-W202-M (WG 11 N8963) Joint Scalable Video Model (JSVM) 10 [07/05/31]

132

50.4.1.1.4 JVT-W203-M (WG 11 N8964) JSVM 10 Software [07/06/29]

50.4.1.1.5 JVT-W204-M (WG 11 N8955) WD 1 conformance test spec for Prof Prof (Teruhiko Suzuki) [07/06/29]

50.4.1.1.6 JVT-W205-M (WG 11 N8957) WD 1 conformance test for SVC (V. Bottreau) [07/06/29]

50.4.1.1.7 JVT-W206-M (WG 11 N8959) WD reference software for Prof Prof [07/06/29]

50.4.1.1.8 JVT-W207-M (WG 11 N8967) Joint Multi-view Video Model (JMVM) 4 [07/05/18]

50.4.1.1.9 JVT-W208-M (WG 11 N8968) JMVM 4 Software [07/05/31]

50.4.1.1.10 JVT-V209-M (WG 11 N8966) Joint draft 3 Multi-view Video Coding [07/02/09]

50.4.1.1.11 JVT-W211-M (WG 11 N8961) WD reference software for SVC [07/06/29]

50.4.1.1.12 JVT-W212-M (WG 11 N8965) Verification test plan for SVC [07/05/18]

50.5 JVT internal output documentsJVT internal output documents included the following. (Dates listed are planned dates of availability.)

50.5.1.1.1 JVT-W210-M ITU-T Rec. H.264 | ISO/IEC 14496-10 Advanced Video Coding Defect Report [07/06/18]

50.6 SVC core experiment output documentsSubmission (to the JVT) of final description (and any data necessary for conducting experiment): next meeting start – 3 weeksSubmission (to the CE partners) of final software and results: next meeting start – 2 weeks

50.6.1.1.1 JVT-W301 CE 1 on SVC subband techiquesCoordinator(s): Shih-Ta HsiangParticipants: Motorola, Nokia, Qualcomm, HHI, Sharp, Microsoft, RWTH Aachen, Thomson,

FT/Orange, Huawei, IntelTechnology to be tested: JVT-W097 (and generalizations discussed – non-dyadic, interlaced, etc.)

50.6.1.1.2 JVT-W302 CE 2 on SVC bit depth and chroma format scalabilityCoordinator(s): Andrew SegallParticipants: Sharp, Thomson, HHI, Qualcomm, Mitsubishi, Microsoft, Intel, Huawei,

Motorola, NTT, ETRITechnology to be tested: JVT-W102, JVT-V078, JVT-W113

133

50.7 MVC core experiment output documentsSubmission (to the JVT) of final description (and any data necessary for conducting experiment): next meeting start – 3 weeksSubmission (to the CE partners) of final software and results: next meeting start – 2 weeks

50.7.1.1.1 JVT-W303 CE 3 on MVC view interpolation/synthesisCoordinator(s): Hideaki KimataParticipants: Nokia, Qualcomm, Thomson, Microsoft, NTT, Samsung, KHU, Sejong Univ.,

KETI, GIST, Yonsei Univ., HHI, Sharp, Mitsubishi, HuaweiTechnology to be tested: JVT-W059 and JVT-W087

51 JVT administrative and liaison topics

51.1 IPR policy reminder and updateParticipants were reminded of the IPR policy established by the parent organizations of the JVT and were referred to the parent body web sites for further information. The IPR policy was summarized for the participants.

Participants were particularly reminded of the need to supply a completed JVT IPR status reporting form in all technical proposals for normative standardization. Participants were also reminded of the need to formally report patent rights to the top-level parent bodies (using the common reporting form found on the database listed below) and to make verbal and/or document IPR reports within the JVT as necessary in the event that they are aware of unreported patents that are essential to implementation of a standard or of a draft standard under development.

The JVT chair noted that the top-level parent bodies have agreed upon a new common patent policy for ITU-T, ITU-R, ISO, and IEC.

Some relevant links for organizational and IPR policy information are provided below:– http://www.itu.int/ITU-T/ipr/index.html (new common patent policy for ITU-T, ITU-R, ISO,

IEC and guidelines and forms for formal reporting to the parent bodies)– http://ftp3.itu.int/av-arch/jvt-site (JVT contribution template for each meeting)– http://www.itu.int/ITU-T/studygroups/com16/jvt/index.html (JVT founding charter)– http://www.itu.int/ITU-T/dbase/patent/index.html (ITU-T IPR database)– http://www.itscj.ipsj.or.jp/sc29/29w7proc.htm (SC29 Procedures)

The JVT chair noted that the ITU TSB director's AHG on IPR had recently issued a clarification of the IPR reporting process for ITU-T standards, as follows (and as previously sent to the JVT email reflector), per upcoming TD 327 (GEN/16):

“TSB has reported to the TSB Director’s IPR Ad Hoc Group that they are receiving Patent Statement and Licensing Declaration forms regarding technology submitted in Contributions that may not yet be incorporated in a draft new or revised Recommendation. The IPR Ad Hoc Group observes that, while disclosure of patent information is strongly encouraged as early as possible, the premature submission of Patent Statement and Licensing Declaration forms is not an appropriate tool for such purpose.

In cases where a contributor wishes to disclose patents related to technology in Contributions, this can be done in the Contributions themselves, or informed verbally or

134

otherwise in written form to the technical group (e.g. a Rapporteur’s group), disclosure which should then be duly noted in the meeting report for future reference and record keeping.

It should be noted that the TSB may not be able to meaningfully classify Patent Statement and Licensing Declaration forms for technology in Contributions, since sometimes there are no means to identify the exact work item to which the disclosure applies, or there is no way to ascertain whether the proposal in a Contribution would be adopted into a draft Recommendation.

Therefore, patent holders should submit the Patent Statement and Licensing Declaration form at the time the patent holder believes that the patent is essential to the implementation of a draft or approved Recommendation.”

The JVT chair noted (as also previously remarked on the JVT email reflector) that since we are nearing completion of the SVC amendment project, it was suggested that now would be a good time to file formal notices to the parent bodies for any patent rights that are believed to be essential to the implementation of the SVC extensions (not to mention any notices not previously filed relating to the new professional profiles or other previous projects).

It is suggested that, to enable proper interpretation of such formal notices, the SVC amendment should be clearly identified in such formal notices. For example, as “ITU-T Rec. H.264 and ISO/IEC 14496-10 Advanced video coding (2005 Ed.) Amendment 3 (2007): Scalable video coding”. Notices pertaining to other efforts should be made with a similar degree of clarity of identification of the specific standardization work item to which the declaration pertains.

The chair invited participants to make any necessary verbal reports of previously-unreported IPR in draft standards under preparation and opened the floor for such reports: No such verbal reports were made.

51.2 Meeting opening remarks by the chairmenAt the opening session of the meeting, the JVT chairs reminded participants of the relevant IPR policy as described above, and reviewed the status and plans for the major projects under way in the JVT, The two largest areas of activity consisting of scalable video coding (SVC) and multi-view video coding (MVC) extensions of the ITU-T Rec. H.264 | ISO/IEC 14496-10 Advanced video coding (AVC) standard. Further work and additional needs on the development, standardization, and maintenance of the base specification and the recently-completed professional profiles, and of associated reference software and conformance specifications was also noted.

The chair remarked that there were fewer late document uploads this time and that the submitted documents seem to be adhering better to the JVT guidelines in terms of formatting, filenames, etc., which is a good development, although further improvement (particularly in the formatting conventions) is still needed. The new JVT operating rules established in Hangzhou that took into effect at the preceding Marrakech meeting on that subject may have helped.

51.3 JVT communication practicesJVT documents are available at http://ftp3.itu.int/av-arch/jvt-site.

135

These can also be accessed via ftp with the site name ftp3.itu.int, user ID avguest and password Avguest. Upon login, documents will then be found in the directory "jvt-site". Uploading of contributions is done by upload via ftp protocol to the "jvt-site/dropbox" directory.

JVT email lists are managed through the site http://mailman.rwth-aachen.de/mailman/options/jvt-xyz, and to send email to one of these reflectors, the email address is "jvt-xyz@lists.rwth-aachen.de", where "xyz" is– "experts" for general experts group discussions– "bitstream" for bitstream exchange activities– "svc" for SVC work– "mvc" for MVC work

51.4 Scheduling and logistics notesSome parallel sessions were held during the meeting, particularly including some parallel review of MVC and SVC contributions (prior to Thursday afternoon). Some “break-out group” (BoG) side activities and informal study efforts were also conducted. Documents produced by break-out group activities are listed in this report with the abbreviation “BoG” and are suffixed with “-B”.

51.5 Closing session notesIn the closing session there were no requests to reopen discussions of preceding agenda topics and side activities recorded elsewhere in this report.

The JVT thanked the USNB to WG 11, and Julie Higgins, Betsy Bartlett and Scott Porter from Meeting Planit for the organization of this meeting.

The JVT also thanked Apple, Microsoft and Mobilygen for providing financial support for the meeting.

The meeting was closed at 1:50 pm on Friday 27 April 2007.

51.6 Administrative documents

51.6.1.1.1 JVT-W001 (Admin) [G. J. Sullivan, J.-R. Ohm, A. Luthra, T. Wiegand] AHG Report: Proj mgmt and errata

This document (available late) is a report of the JVT Ad hoc group on project management and errata reporting. Its purpose is to provide a high-level survey the current state of JVT projects and to report on recent events and progress made since the last meeting. This document’s content consists primarily of a summary of high-level information found also in other inputs to the meeting.

The primary JVT projects, as of our previous meeting, were reported to be the following:– New profiles for professional applications– Scalable video coding (SVC) extensions– Multi-view video coding (MVC) extensionsAdditional smaller efforts, such as definition of new SEI messages for various purposes, were reported to also be under consideration.

Additionally, we have continuing efforts toward:136

– Development of good conformance testing suites– Development of good reference software for:

– Providing guidance to clarify proper decoder results– Providing guidance to ease encoder and decoder product development– Providing examples of encoding algorithms– Providing examples of decoder error/loss concealment behavior– Serving as a “touchstone” for development of future algorithms

– Aiding in verification testing of design capabilities– Identification and aid toward support of JVT video coding specifications in relevant system

designs

JVT communication practices were reviewed and summarized.

Amendment 2, specifying new profiles (designed primarily) for professional applications, was reported to have reached the following status:

– In ITU-T, Amd.2 reached full “in force” Recommendation status (i.e., final standardization approval) on 6 April 2007, but is not yet published. The “last call” period resulted in one set of sector member comments, which were from Microsoft requesting the latest developments in the JVT to be incorporated. The specification draft was changed to address these comments, and was posted for an “additional review” (AR) period. The AR period closed without further comment.

– In ISO/IEC JTC 1, essentially the same Amd.2 draft text as was approved by ITU-T was forwarded to the SC 29 secretariat as an FDIS. It will soon be subject to a final 2-month FDIS approval ballot (the result of which will be either Yes or No – and almost certainly Yes, without changes to the text).

Further relevant information on Amd.2 was reported to be found in JVT-W003.

The SVC Joint Draft (JD 9) and SVC Joint Scalable Video Model (JSVM 9) were reported to have been submitted as JVT-V201 and JVT-V202, respectively.

The editors were reported to have further worked on the JD and JSVM text after providing JVT-V202. Updated versions of the texts were reported to have been provided as input document JVT-W070.

The JSVM 8 software was reported to have been delivered to the group at the end of the Marrakech meeting. The JSVM software integration process was reported to have followed the rules and procedures defined in the JSVM Software Manual available in the CVS server.

CVS reference:host address: garcon.ient.rwth-aachen.de user name: jvtuser password: jvt.Amd.2authentication: pserver path: /cvs/jvt module name: jsvm_red

Some integration work on text and software was reported to remain to be finalized.

Four SVC “core experiments” were reported to have been the subject of work since the Marrakech meeting. Some of these are toward topics for the current first phase of SVC work and some are for subjects identified as longer term “phase 2” study efforts, as follows:– CE 1: FGS simplification (phase 2)

137

– CE 2: ESS improvement (phase 1)– CE 3: Subband intra coding (phase 2)– CE 4: Bit depth scalability (phase 2)Input reports of work on these experiments have been provided as input contributions to this meeting.

Further relevant information on SVC work was reported to be found in JVT-W006 (and JVT-W007, JVT-W008, JVT-W009, and JVT-W010).

The JMVM 3 and MVC JD 2 were reported to have been submitted to the JVT as JVT-V207 and JVT-V209, respectively.

The JMVM 3 software was reported to have been delivered to the group on February 24th, 2007. This release was reported to contain the integration of new syntax element as described in JVT-V054, reference list reordering commands for inter-view pictures as described in JVT-V043, bug fixes and code clean-ups. Subsequently two bug-fix versions tagged JMVM 3_0_1 and JMVM_3_0_2 were reported to have been released which contained significant bug-fixes which addressed the high memory usage and spatial direct mode.

CVS reference:host address: garcon.ient.rwth-aachen.deuser name: jvtuser password: jvt.Amd.2authentication: pserver path: /cvs/jvt module name: jmvm or jmvm_red

jmvm_red does not check out certain old folders related to SVC.

Two MVC “core experiments” were reported to have been the subject of work since the Marrakech meeting, as follows:– CE 5: Illumination compensation– CE 6: View interpolationInput reports of work on these experiments have been provided as input contributions to this meeting.

Further relevant information on MVC work was reported to be found in JVT-W013 (and JVT-W011, JVT-W012, and JVT-W014).

The latest available state of errata reporting on the AVC base specification was reported to be found in JVT-U210, plus relevant notes in the meeting report of the Marrakech meeting. The San Jose input document JVT-W134 was also reported to be relevant.

As of the writing of the report, the latest errata list JVT-V210 (planned as a JVT internal output document in Marrakech), had not yet been produced. Hope was expressed for it, or a further-updated errata list to be produced as an output document from the San Jose meeting, to be produced soon.

The latest JM algorithm description text was reported to have been submitted as JVT-W057.

JM software versions 12.1 and 12.2 were reported to have been released since the Marrakech meeting

Improvements to the JM software are described in JVT-W044.

138

The integration of the new 4:4:4 profiles had reportedly been started and was still a work in progress.

The software and updated documentation is available at:http://iphome.hhi.de/suehring/tml

The JM software manual had reportedly been updated to match the released version JM 12.2 and had been submitted to this meeting as document JVT-W041.

A web based bug tracking system had been set up for keeping track of known issues and missing features. The system is publicly accessible but requires registration for entering bug reports.

The system is located athttp://ipbt.hhi.de

A list of known issues and their state can be found at:https://ipbt.hhi.de/mantis/view_all_bug_page.php

Further relevant information was reported to be found in JVT-W002.

The JVT, as a child organization with parents in ISO/IEC JTC 1 and ITU-T, is operated under the top-level IPR policies of these organizations. Two recent noteworthy developments were reported to have occurred in the IPR policies of these top-level organizations.1) The top-level parent bodies have agreed upon a new common patent policy for ITU-T, ITU-

R, ISO, and IEC. That policy, and guidelines and forms for formal reporting to the parent bodies, can be found at http://www.itu.int/ITU-T/ipr/index.html.

2) The ITU TSB director's AHG on IPR had recently issued a clarification of the IPR reporting process for ITU-T standards (as previously sent to the JVT email reflector), per upcoming ITU-T TD 327 (GEN/16).

51.6.1.1.2 JVT-W002 (Admin) [K. Suehring, A. Tourapis, T. Suzuki] AHG Report: JM text, ref soft, bitstream, conf

The JM reference text includes the adopted contribution of the document JVT-T046 on "Context Adaptive Lagrange Multiplier (CALM) for Motion Estimation in JM - Improvement". It had been submitted as document JVT-W057.

The integration of the 4:4:4 profiles had been started and was still work in progress. JM 12.2 (see Software releases) already contained the code for of Tone Mapping and Post Filter hint SEI messages as well as the Intra only profiles. The Independent Color Coding mode software had been finished, but the code had not yet been released.

The JM versions 12.1 and 12.2 had been released since the Marrakech meeting. Besides the 4:4:4 features, the main focus of these releases was restructuring, code improvement and speedup. The decoder runs at least more than twice as fast than previous versions.

The most important improvements are described in JVT-W044.

The complete list of changes can be found in the CHANGES.TXT file which is included in each software archive.

The software and updated documentation is available at:http://iphome.hhi.de/suehring/tml

139

The JM software manual had been updated to match the released version JM 12.2 and had been submitted to this meeting as document JVT-W041. It was reportedly planned to add the manual to the software archive in subsequent versions.

As the official H.264/AVC reference software, the JM should be a correct source for checking implementations. This means the decoder should be able to decode all valid H.264/AVC bitstreams and the encoder should never create non-conforming bitstreams (at least not without generating warnings). This is currently not the case.

Depending on the configuration the JM encoder can create invalid bitstreams:– Level constraints are not properly checked– The 16-bit transform processing range requirements are not checked– In Baseline/Main/Extended profile the restriction of CAVLC syntax elements needs proper

handling

The software coordinators encouraged all H.264/AVC experts to volunteer for fixing these issues.

A web based bug tracking system has been set up for keeping track of known issues and missing features. The system is publicly accessible but requires registration for entering bug reports.

The system is located athttp://ipbt.hhi.de

This internet site contains some usage instructions.

Please note that the bug tracking system is using encrypted/secure http (https) for protecting the user’s login. The used certificate is self signed and has to be imported into the user’s web browser. The SHA-1 fingerprint of the certificate is

69:21:86:d9:3e:72:da:3f:e8:30:df:a8:dd:fa:a5:4c:ed:85:b5:09

A list of known issues and their state can be found at:https://ipbt.hhi.de/mantis/view_all_bug_page.php

A list of current bugs can also be found in the annex of the AHG report.

It was requested that certain rules should be followed before reporting any new bugs:– The database should be searched on whether the same issue was previously reported. If the

problem was reported before, but there is additional information, then this information should be added to the original report.

– It should be specified if the problem is related to the encoder, decoder or both.– The version of the software used should be specified.– Description of the problem should be as precise as possible.– The necessary steps to reproduce the problem should be described in detail.– If available, the configuration files or/and command line syntax used to run the software

should be provided.– The language of the standard should be used when referencing the text description.– After filing the report, the user should check if he/she is requested to provide additional or

other information relating to this issue.

Communications related to this ad-hoc activity have taken place on the JVT bitstream exchange reflector (“jvt-bitstream@lists.rwth-aachen.de”). The reflector of this AHG was moved from

140

IMTC to Univ. of Aachen some time ago. However AHG was not so active since the last JVT meeting.

The FTP area for downloading bitstream files is on the main JVT Experts FTP site:ftp://ftp3.itu.int/jvt-site/bitstream_exchange/ (login: avguest, password Avguest).

The bitstreams can also be accessed from the following http site.http://ftp3.itu.int/av-arch/jvt-site/bitstream_exchange/

To volunteer a bitstream for testing, please include it in a zip archive along with related files (trace files, configuration, reconstructed frames) in a zip archive and upload it to the dropbox:

ftp://ftp3.itu.int/jvt-site/dropbox (login: avguest, password Avguest)

In general, the following naming convention is being followed for the bitstreams in the exchange:

FeatureCode_Source_VersionLetter

Please refer to the spreadsheet and files on the FTP site for examples.

Once a bitstream has been uploaded to the dropbox, send an e-mail to teruhiko@av.crl.sony.co.jp, and/or the bitstream exchange reflector and it will be made available in the bitstream_exchange directory.

To sign up for the bitstream exchange reflector, use the web address given below. – Over the web: < http://mailman.rwth-aachen.de/mailman/listinfo/jvt-bitstream >

Conformance Activities and Corrigendum work:No new conformance specification problems were reported since the last meeting. All known problems must be fixed for the corrigendum of AVC conformance and FRExt conformance.

New conformance activity for new professional profiles should be started at San Jose meeting.

The AHG recommended– to fix all bitstreams with conformance problems– to encourage volunteers to provide more conformance streams– to start activity of new conformance amendment work to support new professional profiles

A desire for corrigendum and software work was expressed – e.g., range of values checking and avoiding allowing “hostile” non-conforming bitstream corner cases.

51.6.1.1.3 JVT-W003 (Admin) [T. Suzuki] AHG Report: Professional applicationsThe main JVT reflector (jvt-experts@lists.rwth-aachen.de) was used for the AHG activities. The term “[4:4:4]” was inserted at the beginning of a subject field to identify the email related to this AHG. The descriptions of the specifications were updated, e.g. description of independent color mode, and the FDAM document was released to ISO and the corresponding AR text was released to ITU-T. All remaining issues on the document were fixed in those documents. The remaining issues are the reference software and conformance.

The volunteers to integrate JFVM software into the latest JM were identified. The order of integration is as follows.

1) Tone Mapping SEI (Sharp)2) 4:2:0/4:2:2intra only coding & post-filter hint SEI (Panasonic)

141

3) independent color coding (Mitsubishi)4) 4:4:4 intra & predictive (Thomson)5) Lossless coding (Sejong University)

The integration had not been finished by this meeting. The integration of 1), 2) and 3) had been finished, however the schedule was delayed. 4) and 5) should be integrated after the San Jose meeting. The schedule of formal integration remained to be defined during San Jose meeting.

Conformance streams for new profiles will be started after San Jose meeting. The following volunteers were identified. High 4:4:4 Thomson Mitsubishi (independent color coding mode) Sejong Univ. (lossless coding)High 4:4:4 intra Thomson Mitsubishi (independent color coding mode) Sejong Univ. (lossless coding)CAVLC 4:4:4 intra ThomsonHigh 4:2:2 intra PanasonicHigh 10 intra Panasonic

It was encouraged to generate bitstreams by other volunteers in addition to the above organizations. The schedule of conformance work plan was recommended to be defined at San Jose.

The AHG recommended– To finalize the work plan of the integration of JM software– To finalize the conformance work plan

Various identified problems have been fixed.

Regarding the software – separate color plane coding (Thomson) and lossless coding reamained still to be done.

Regarding conformance – volunteers were listed – additional volunteers would be helpful, should also tabulate further detail.

51.6.1.1.4 JVT-W004 (Admin) [J.-R. Ohm, T. Wiegand, M. Bober] AHG Report: Video annotation

During the interim period since the last JVT meeting, some active email discussions of video annotation were held on the JVT email reflector. These consisted primarily of an airing of views regarding where it is best to carry video annotation data (i.e., at the systems level or within the video bitstream as SEI messages or registered or unregistered user data SEI messages), where it is best to specify the definition of such data (i.e., in a separate standard such as the MPEG-7 standard or in particular SEI message definition sections of the AVC standard), and how to deal with an asserted confusion resulting from an asserted overabundance of defined types of such data. Various views were expressed, along with pros and cons of each approach. No obvious consensus was evident on those issues.

Various perspectives were expressed, ranging from doing nothing (letting people use user data SEI or system level support) to selecting particular messages for definition in SEI. Discussion of scope, system interaction, specification interaction, …

142

The JVT-W032, JVT-W033, and JVT-W034 input contributions to the San Jose meeting are relevant to the subject.

51.6.1.1.5 JVT-W005 (Admin) [G. J. Sullivan, J. Luo] AHG Report: AVC splicingEmail to initiate the discussion was sent to the JVT reflector. But little discussion occurred there. The normative requirements may often make concatenation/splicing of coded video sequences rather difficult.

There is one AVC HRD related proposal in the San Jose meeting (JVT-W020). It may have some relevance. The study of the issues should be continued and action items should be identified. Contributions are needed to determine what can be done.

Basically, not much happened in this AHG.

51.6.1.1.6 JVT-W006 (Admin) [J. Vieron, M. Wien, H. Schwarz, T. Wiegand] AHG Report: JD & JSVM text and S/W

The SVC Joint Draft (JD 9) and SVC Joint Scalable Video Model (JSVM 9) were reported to have been submitted as JVT-V201 and JVT-V202, respectively. They were also reported to have been submitted as MPEG output documents N8750 (Study Text of ISO/IEC 14496-10:2005/FPDAM3 Scalable Video Coding) and N8751 (Joint Scalable Video Model (JSVM) 9). The provided Joint Draft 9 corresponds to JSVM 8 Annex G with FGS removed.

The JSVM 9 document includes a generic description of the principles used for scalable coding in SVC to help people to get familiar with scalability principles. It also includes a description of non-normative tools for the encoding process.

The JSVM 9 also includes an annex (Annex-G) corresponding to a modified version of the JD 9 including all tools adopted during the 22nd JVT meeting. The purpose of this additional document is to serve as a base for the creation of the future JD 10.

The document JVT-V202_JSVM9.doc contains a new part (Annex A), that contains the draft text for FGS, which was removed from the Joint Draft 9, including dedicated subclauses and the specification of changes to subclauses in Annex G that are required for application of FGS in SVC.

Presented. FGS moved to “Annex A”. Lots of work. Software integration not done. Not much feedback form members on text. Feedback requested. Members requested to strictly respect the rules and procedures.

Normative changes– Moving FGS and AR-FGS from JD back to JSVM (not Annex G) [Editors]– JVT-V032* [J. He] CE4: Disabling SVC chroma deblocking filter (as values of

disable_deblocking_idc)– JVT-V035* [A. Segall] CE8: CGS SVC-to-AVC bitstream-rewriting (incl removal of IDCT

for base layer of MGS/CGS SNR scalability)– Remove the use of nal_unit_type value 21, using spare bits in the current use of

nal_unit_type equal to 20 [Editors]– constrained_intra_pred_flag must be 1 when Intra_base is used [Editors]– disallow temporal direct for nal_unit_type = 20 or 21 [Editors]

143

– When nal_unit_type = 1,2,3,4, then disallow temporal direct when used for inter-layer prediction [Editors]

– Number of base layer macroblocks that need to be decoded in order to form an IntraBL predictor should be limited. [T. Wiegand, details TBD]

– suffix NAL unit – nal_ref_idc must be the same as the associated non-suffix NAL units [Y.-K. Wang]

– semantics of discardable_flag – is an indication of a lack of dependency for the current access unit and all subsequent access units [Y.-K. Wang]

– Bitstream may require discarding of some NAL units with simple_priority_id … in order to form a conforming subset bitstream [Y.-K. Wang]

– Smoothed reference prediction flag to be at slice level rather than in SPS [N. Cammas]– Put a flag in the slice header to “skip” the entire slice (indicating that BLSkip flag is equal to

1 and residual_pred-flag is equal to 1 for all MBs in slice and no further information is sent) [Editors]

– JVT-V068* [J. Luo] SVC hypothetical reference decoder (Details as recorded elsewhere), r4 SEI message for temporal subsets and other aspects

– JVT-V088* [A. Eleftheriadis] SVC error resil using frame index in NAL unit header (extra byte for any D, Q; byte is moved from NAL unit header to slice header and suffix/prefix NAL unit payload; flag for switching the byte stays in the NAL unit header)

– Suffix NAL unit – suggested change removes the ability to provide more than one of these. Alternative suggestion – for the first NAL unit of the base layer, a NAL unit type 14 is prefixed to convey the contents of the current suffix NAL unit. For other slices, we use the suffix NAL units as they are (NAL unit type 20). [Y.-K. Wang]

– Prohibit MV refinement when store_base_layer_flag = 1 [Editors]– JVT-V036* [A. Segall] Support for transcoding in scalability info SEI (r1)– SVC profile and levels according to meeting notes (incl removal of profile C) [Editors]

For FGS (integrated into JVT-W070)– JVT-V095* [M. Karczewicz] CE1: Improved coefficient coding (Tool 1: adopted; Tool 2:

change only for I and P but not B slices)

Issues (JD)– Feedback provided by JVT members on the documents was very low. Comments received

from Nokia (Ye-Kui) and Microsoft (Gary)– Clarify constraints for frame_num (inside a “layer”)– Clarify usage of frame_num for base layers– Move process 8.2.2 to Clause 7 (slice groups)– Rewrite resampling G.8.6 (highly redundant)– Ed. Notes to be solved– (anything else that’s broken, needs to be clarified)

Issues (JSVM)– FGS part needs to be reworked and improved

The editors further worked on the JD and JSVM text after providing JVT-V202. Updated versions of the texts are provided as input document JVT-W070. They contain various changes including the following:JSVM:– Integration of JVT-V095 (was missing in JVT-V202)– JD:– Add/extend definitions in G.3

144

– Corrections, clarifications in G.6– Update of G.7 (syntax and semantics) + several fixes– Restructuring of G.8– Several corrections, clarifications in G.8– Correction of bugs in G.10 (profiles and levels)

The JSVM 8 software was delivered to the group at the end of the Marrakech meeting. The JSVM software integration process has followed the rules and procedures defined in the JSVM Software Manual available in the CVS server listed below.

Note that the integration process has more than one month delay with respect to the original integration schedule. Various integrators encountered difficulties which were reportedly mainly related to the FGS parts. There were still four software integrations to be done.

The last JSVM integration schedule is summarized in the table below.Proposal Company Start date Comments JSVM Tag Status

Cleaning + Memory leaks fixing + Improvement of FixedQPEncoder

Thomson 01.02.2007 simple(1 day) JSVM_8_0_1 OK

[JVT-T037] - CE2: Progressive to interlace inter layer motion prediction

Samsung 01.02.2007 Moderate(2 days) JSVM_8_1 OK

[JVT-V088] - tl0_picture_idx Layered Media 07.02.2007 Simple (2 days) JSVM_8_2 OK

[JVT-V125]- H241 RCDO HHI 12.02.2007 Moderate (3 days ) JSVM_8_3 OK

[JVT-V126] MGS Key pictures HHI 15.02.2007 Moderate (3 days ) JSVM_8_4 OK

[JVT-V074][JVT-V090] - Motion comp interpolation 4-tap and parameterized

Microsoft 18.02.2007 Moderate (4 days) JSVM_8_5 OK

[JVT-V058] - Smoothed reference flag + Interlaced Bug fixes (J. Vieron)

Orange 26.02.2007 Simple (1 day) JSVM_8_6 OK

[JVT-V032] - CE4: Disabling SVC chroma deblocking filter

Freescale 27.02.2007 Moderate(5 days) JSVM_8_7 OK

[JVT-V035]– Bitstream rewriting ([JVT-V036])

Sharp 02.03.2007 Moderate (7 days) JSVM_8_8 OK

High-level syntax Nokia 16.03.2007 Moderate (4 days) JSVM_8_9 OK

CGS/MGS residual prediction in transform domian Sharp 20.03.2007 Difficult JSVM_8_X Started

[JVT-V068] - HRD and SEI message Thomson Difficult / parallel JSVM_8_X

[JVT-V079] - SVC low complexity MB mode decision ST Microelectronics Moderate (7 days) JSVM_8_X

FGS refinements Qualcomm Moderate (4 days) JSVM_8_XEnd of JSVM 9 integration JSVM_9_0

In order to improve the whole software integration process, the software integration guidelines and rules have been refined as following:– The integrated software shall compile without warnings when using the provided VS 6,

VS .NET, and VS 2005 workspaces, as well as linux makefiles.– Do not use variable declarations inside the header of for-loops (the scope for for-loops is not

correctly supported with all compilers).– Follow the coding style of the JSVM software. Use 2 (two) spaces for indentation, no tab.– Re-use code and integrate functionality as possible. Try to avoid redundant code.– Do not change the meaning of existing input parameters but define new ones if necessary

(and applicable). – Make sure that new parameters have meaningful default values. Tools should not be

switched on by default (if not decided different by the JVT).– Do not re-structure the output of the compiled binaries (if not decided different by the JVT).

145

– Please change the JSVM version number macro (i.e. “_JSVM_VERSION_”) located in the file “CommonDefs.h” to be inline with your integration tag.

The AhG on SVC text and JSVM software recommended1) To take the proposed version of the JD and JSVM text in JVT-W070 as basis for further

editing.2) To carefully study the latest version of the JD (JVT-W070) and provide feedback to the

editors.3) To follow the integration rules and procedure of validation described in the JSVM software

manual (found on the CVS server listed below).4) All proponents to strictly respect these rules/guidelines. Sticking to these basic principles and

recommendations is mandatory and facilitate the future integration and maintenance works.5) To continue maintaining the JSVM Software Manual: We remind that each proponent is

responsible for updating the Software Manual by providing description for each newly introduced parameters and/or tools.

CVS reference:host address: garcon.ient.rwth-aachen.de user name: jvtuser password: jvt.Amd.2authentication: pserver path: /cvs/jvt module name: jsvm_red

51.6.1.1.7 JVT-W007 (Admin) [A. Segall, S. Regunathan] AHG Report: SS resampling

A kick-off message was sent to the JVT reflector on 27 February 2007. The message requested suggestions on upsamplers, down-samplers and sequences with "different source characteristics" from interested experts.

Candidate down-samplers were identified for study and circulated on the reflector on 4 April 2007. The AhG decided to study filters from the paper: K. Turkowski “Filters for Common Resampling Tasks”. (Online at http://www.worldserver.com/turk/computergraphics/ResamplingFilters.pdf) The AhG recommended studying the two Gaussian filters and two Lanczos windowed filters described in the document. Furthermore, the AhG recommended combining the Gaussian/Lanczos filters with an unsharp mask to better approximate common image enhancement techniques. Regarding the sample ratios, the AhG recommended interested experts to focus on the case of dyadic, 1.25 and 1.5 cases. This was to cover common NTSC->HD and PAL->HD (accounting for the change in aspect ratio), as well as 720p->1080p and 1080i->1080p applications.

The AhG conducted an internal evaluation of the proposed down-samplers. The AhG did not find evidence that additional upsampling filters are needed in the SVC specification.

Documents JVT-W022 and JVT-W028 relate to filter design. Focus was`on dyadic, 1.25, and 1.5. A number of other documents were listed as relevant.

CE2 has a residual upsampling part.CE3 is related to subband spatial scalability.

Relevant contributions:– Resampling: JVT-W022, JVT-W028, JVT-W086– Spatial scalability: JVT-W097, JVT-W122

146

– Inter-layer prediction: JVT-W105, JVT-W109, JVT-W117, JVT-W106, JVT-W130, JVT-W123

51.6.1.1.8 JVT-W008 (Admin) [H. Schwarz, S. Regunathan, A. Eleftheriadis] AHG Report: SVC complexity reduction

Relevant contributions were listed and summarized, including: JVT-W027, JVT-W029, JVT-W061, JVT-W063, JVT-W068, JVT-W069, and JVT-W072.

51.6.1.1.9 JVT-W009 (Admin) [Y.-K. Wang, S. Pateux, P. Amon, T. Schierl] AHG Report: SVC high-level syntax, err resil

There have been some discussions regarding signaling of full sets of HRD parameters for rewritten bitstreams. One counter argument was reportedly that signaling full sets of HRD parameters for rewritten bitstreams is somehow overkill, because SVC already supports signaling of one full set of HRD parameters for each operation point or scalable layer, and SVC has not yet supported signaling of HRD parameters for extracted bitstreams according to quality layer information or priority_id values. JVT-W091 was reported to be related to this topic.

Nokia and University of Science and Technology of China were reported to have started the following implementation work to the JSVM.– Coding of multiple slices per picture– Slice size of fixed number of macroblocks– Slice size of fixed number of bytesThe implementation for slice size of fixed number of macroblocks was reported to have been finished, and the other part was reported as ongoing.

Relevant contributions were listed (High-level syntax: JVT-W020, JVT-W046, JVT-W047, JVT-W048, JVT-W051, JVT-W052, JVT-W053, JVT-W064, JVT-W091, JVT-W114, and JVT-W125; Error resilience: JVT-W049, JVT-W050, JVT-W054, and JVT-W062).

51.6.1.1.10 JVT-W010 (Admin) [Y. Gao, A. Segall, T. Wiegand] AHG Report: SVC bit depth and chroma format

Relevant contributions and the status of work on CE4 were noted.

The AhG sent a kick-off message to the JVT main reflector (jvt-experts@lists.rwth-aachen.de) on 7 March 2007. There were no other messages on the reflector. The work of the AhG consisted of generating test conditions and test sequences for CE4. Test conditions were circulated by the AhG in the kick-off message and utilized for testing within CE4. Sequences were generated within CE4 by the CE partners. The procedure for generating the test sequences was provided in an Appendix to the AHG report, and includes representative tone mapping and linear shifting operations.

Test material discussion – test sequences were generated for tone mapping and linear shifting. Used for CE4. An appendix to the report describes the creation process for test sequences for bit-depth scalability used in CE4.

The following contributions were noted in the AHG report:– JVT-W102 and JVT-W113 on bit depth scalability– JVT-W076 on chroma format scalability – there may be an issue with the notion of SNR

scalability happening in luma while spatial scalability is happening in the chroma. Design is 147

OK as long as luma spatial scalability is happening whenever chroma spatial scalability is happening.

51.6.1.1.11 JVT-W011 (Admin) [A. Vetro, P. Pandit] AHG Report: MVC high-level syntax & buffering

There was reportedly some reflector discussion on the subject of SPS and base views since the last meeting. Based on this discussion, it was reported that some issues might need some further discussion and possible clarification in the text, including the following.– Need to further clarify and confirm the differences between IDR pictures, View-IDR

pictures, and anchor pictures? For an anchor picture, pictures later in decoding order but earlier in output order than the anchor picture may refer to pictures earlier in decoding order than the anchor picture. For an IDR picture, no picture later in decoding order than the IDR picture may refer to pictures earlier in decoding order than the anchor picture.

– Need to clarify the marking of pictures as unused for reference. In AVC, for an IDR picture, all previous pictures in decoding order are marked as "unused for reference". In JD2.0 of MVC, V-IDR does this for a view.

– Is the SPS allowed to change at a P or B-picture? According to the current spec, the SPS shall be changed in an IDR access unit only. However, this might need some additional clarification in the MVC context since an access unit with an IDR picture might contain P or B-pictures.

– How is the view_id for base view indicated? If the base-view is an independently decodable view with NAL unit type 20, then NAL unit header includes the view_id. If the base-view is an AVC-compatible view, then the prefix/suffix NAL unit will carry the view_id information for that view, which the MVC decoder can decode.

51.6.1.1.12 JVT-W012 (Admin) [H.-S. Koo] AHG Report: MVC motion/disparity vector coding

Relevant contributions to the meeting were classified into 3 categories:– new inter prediction process,– modification of motion vector predictor, and– modification of spatial direct mode.Results under common testing conditions: 0.18 dB average / 0.54 dB best case for category 1, 0.0x dB for category 2&3 combined.

The relationship with reference picture list reordering (RPLR) was reportedly being investigated.

51.6.1.1.13 JVT-W013 (Admin) [H. Kimata, A. Smolic, P. Pandit, A. Vetro, C. Ying] AHG Report: JMVM & JD text editing

The JMVM3 and JD2 were submitted to JVT as JVT-V207 and JVT-V209, respectively. Text for an SEI message for parallel processing was added. The JD text included high-level syntax and decoding processes related to reference picture list reordering, text corresponding to the hypothetical reference decoder for MVC and view coding order information in SPS.

Several other editorial improvements and clarifications had also been made to the JD and JMVM text. For the JD text, some minor updates to the SPS semantics had been made and the document included a revised definition of access unit that was reported to be in line with the latest versions of the various AVC amendments. For the JMVM text, the SEI message on parallel processing had been updated as well. These revisions should be considered as editor’s input to the meeting and were included as an attachment to this AHG report.

148

Further issues that had been raised related to high-level syntax and buffering were reported in another AhG report: JVT-W011.

The JMVM 3 software was delivered to the group on February 24th, 2007. This release contained the integration of new syntax element as described in JVT-V054, reference list reordering commands for inter-view pictures as described in JVT-V043, bug fixes and code clean-ups. Subsequently two bug-fix versions tagged JMVM 3_0_1 and JMVM_3_0_2 were released which contained significant bug-fixes which addressed the high memory usage and spatial direct mode.

The work on the software completed so far was summarized as follows:– Add new syntax view_id in SPS to indicate view coding order– Send reference list information in view coding order– Reference Picture List Construction for MVC, including new RPLR (JVT-V043)– Memory reduction for the decoder: remove useless code related to FGS, MCTF and save the

memory to around 1/4– Effective DPB allocation at the encoder – Bug fix for spatial direct mode– Encoder parameter file to read multiple inter-view ref– Some code cleanup for software improvement.

Some software issues that were reported to still need to be addressed were:– Disabled co-located condition for inter-view (limitation of s/w) – An AVC compatible SPS needed to decode AVC compatible view only– Output order of views is not sequential or parallel. It is on an as ready basis.– All the macros need to be cleaned up & removed permanently– Encoder/decoder trace file for each view needed

The manual had been added as part of the JMVM reference software module.

The AhG on JMVM and JD text editing recommended:– To consider the editor’s input in preparing future versions of the JMVM and JD.– To discuss the issues in the current version of the software as mentioned above– To improve the manual created for the JMVM software– To follow the same software integration guidelines present in JSVM (repeated below)

In order to improve the whole software integration process, the software integration guidelines and rules were reported to be as follows:– The integrated software shall compile without warnings when using the provided VC6 and,

VS .NET workspaces, as well as linux makefiles.– Do not use variable declarations inside the header of for-loops (the scope for for-loops is not

correctly supported with all compilers).– Follow the coding style of the JMVM software. Use 2 (two) spaces for indentation, no tabs.– Re-use code and integrate functionality as possible. Try to avoid redundant code.– Do not change the meaning of existing input parameters but define new ones if necessary

(and applicable).– Make sure that new parameters have meaningful default values. Tools should not be

switched on by default (if not decided different by the JVT).– Do not re-structure the output of the compiled binaries (if not decided different by the JVT).– Please change the JMVM version number macro (i.e. “_JMVM_VERSION_”) located in the

file “CommonDefs.h” to be in line with your integration tag.149

CVS reference:host address: garcon.ient.rwth-aachen.deuser name: jvtuser password: jvt.Amd.2authentication: pserver path: /cvs/jvt module name: jmvm or jmvm_red

jmvm_red does not check out certain old folders related to SVC.

The report included a proposed update of text with clarifications, also software updates & bug fixes were proposed. A plan to install a bug reporting system for the software was described.

51.6.1.1.14 JVT-W014 (Admin) [H. Kimata, A. Smolic] AHG Report: MVC exper. framework & test cond

Some discussions on the subjects of “combination of MVC and SVC” and “Multi-view Video plus Depth” were made. Especially for the first topic, spatial scalability in MVC was discussed. These discussions were to initiate new directions of MVC.

Discussions led on reflector, several input contributions, no conclusion yet.

The AHG on MVC experimental framework and testing conditions recommended discussing these new directions of MVC based on relevant input contributions.

51.7 JVT liaison communications

51.7.1.1.1 M14548 WG 11 input [FLO Forum] Liaison statement from FLO Forum to WG 11

M14548 from FLO Forum to WG 11 was noted – It reports the adoption of ISO/IEC 14496-10 / ITU-T H.264 (AVC) Extended Profile Level 1.3 for use in MediaFlo systems in terrestrial mobile multimedia multicast networks. WG 11 (MPEG) is planning to reply to it.

52 Scalable video coding

52.1 CE 1 & related docs: SVC FGS simplification

52.1.1.1.1 JVT-W090 ( Prop 2.2/3.1) [H. Kirchhoffer, H. Schwarz, T. Wiegand] CE1: Simplified FGS

This contribution describes a modification of the transform coefficient level coding of non-PR-slices in SVC. A range of scan positions is specified in the slice header that defines which of the 16 transform coefficient level scan positions of each block (in zig-zag-scan order) is encoded in this slice. In this way, it is possible to divide the transform coefficient levels of an arbitrary non-PR-slice to multiple additional MGS slices and to achieve fine granular SNR scalability. The complexity increase depends on the number of additional MGS layers used and is thus controllable by the encoder.

Idea is enhancing non-PR slices to achieve FGS functionality using MGS. Send a start index and an end index for coefficient frequencies in a slice. Suggest control of complexity by profile/level constraints.

Experiments did not alter encoding rules.

150

No spatial intra modes in enhancement layer – some issues in current software. Interaction with notion of not using spatial domain for SNR scalability.

Currently available software doesn’t yet implement transform-domain prediction.

Inter-layer prediction needs clarification regarding intra prediction processing – concept is workable but decoder is complex. Some options:– Disallow spatially-predicted Intra in enhancement layers (seldom selected anyway –

typically IntraBL is used)– Treat IntraBL neighbor as Inter (not available) for purposes of constrained intra prediction.JVT decision: 2nd approach is adopted.

How to finalize other issue from last meeting: “Number of base layer macroblocks that need to be decoded in order to form an IntraBL predictor should be limited. [T. Wiegand, details TBD]” – some details may need finishing. Addressed in JVT-W070 – intra MBs in base layer that are required for decoding the enhancement layer shall not exceed the number of IntraBL macroblocks in the base layer times 1.5. JVT decision: Agreed.

Used QP difference of 6, two MGS layers. Coefficients that are received are added to the ones previously received.

Some mismatch using current software, but basic concept seems understood and verified. Without encoder-decoder mismatch problem, there would be no difference in the PSNR of the high bit rate point – only a difference in bit rate.

Complexity? This is a small change to decoding parsing process – very small impact.

JVT decision: Adopted.

52.1.1.1.2 JVT-W115 -QV (Late Info) [A. Segall] CE1: Verif JVT-W090 simplified FGS

This document reports a verification of JVT-W090. The proponents provided Sharp with source code and simulations results. Sharp inspected the source code and reported that it confirmed that it matched the proposal. Additionally, Sharp compiled the source code, re-generated the results reported in JVT-W090, and randomly checked data points between results generated at Sharp and provided by the proponent. All checks matched, and the results in JVT-W090 were reported to have been verified.

Verified using provided source code. All checked sequences matched.

52.1.1.1.3 JVT-W111 ( Prop 2.2) [M. Karczewicz, S. Park, H. Chung] CE1: Report on FGS simplif

This contribution proposes changes to the FGS joint significant and refinement coefficient coding method described in JVT-V077. The results reportedly indicate that the joint significant and refinement coefficient coding does not degrade the performance – the average improvement on all tested CIF sequences is reported to be 0.46% and 4CIF sequences to be 0.7%.

JVT-V077/JVT-W121 with simplified sign coding.

Remark: Suggestion to have proponents of JVT-W111 and JVT-W121 confer and report back.

151

Further discussion then held on Thursday. Merged proposal presented as JVT-W121r1. Merged proposal (upload as rev of 121 doc). No penalty at first FGS layer, average penalty goes up to 0.4% for higher FGS layers. JVT decision: Adopted (to FGS part of JSVM).

52.1.1.1.4 JVT-W124 -QV (Late Info) [J. Ridge] CE1: Verif JVT-W111 FGS simplifThe results presented in JVT-W111 were verified and found to be correct.

52.1.1.1.5 JVT-W121 ( Prop 2.2.1/3.1) [J. Ridge, X. Wang] CE1: FGS refinement pass simplif

This contribution proposes to include FGS refinement pass coefficients in the run-length codes previously only associated with significance pass coefficients. While the distinction between “significance” and “refinement” coefficients would remain, there would no longer be a distinct “significance pass” and “refinement pass”. Sign bits for non-zero refinement values for a block would be grouped and transmitted after the end-of-block is reached. It is claimed that this proposal would simplify the FGS VLC algorithm, both in terms of specification and implementation, because there would be no need for two different coding algorithms for significance and refinement passes and because coefficients would be decoded in sequence. An average coding penalty of 0.4% bit rate is reported to be associated with this proposal for QCIF and CIF sequences.

Same as prior JVT-V077. See notes in section on JVT-W111.

52.1.1.1.6 JVT-W119 ( Prop 2.0/3.1) [Y. Bao, M. Karczewicz, X. Wang, J. Ridge, Y. Ye, W. J. Han, S. Y. Kim] CE1: FGS simplif

This contribution reports results of CE1 on FGS simplification to address the concerns on FGS complexity. This contribution proposes to align FGS layer coding with a H.264/AVC baseline base layer and make decoding process with Cycle Aligned Fragment mandatory to reduce the computation complexity and simplify the FGS specification. These two changes along with other simplifications reportedly make it possible to reduce the FGS text to around 30 pages.

Text editing and other simplification of FGS. Several changes discussed and evaluated. Some doubt expressed about AR-FGS aspect – results not yet available to confirm the simplification.

Aspects seem generally agreed.

JVT decision: Adopted (into FGS JSVM “Annex A”, which is an ongoing study item; 90 pages 43 pages, which still includes about 10 or 11 pages of duplicated stuff for context).

52.1.1.1.7 JVT-W120 ( Info) [P. Yin] CE1: Verif JVT-W119 FGS simplifThis contribution reports cross-check result for the proposal by Qualcomm as described in document JVT-W119 “CE1 report: FGS simplification”. The source code and configuration files were provided by QualComm. The provided source code was compiled and the encoder and decoder executable were run with the provided configuration files. All results in terms of R-D were reportedly the same as those provided by QualComm. The decoder crashed for Crew 4CIF at one point.

Verifies JVT-W119.

152

52.2 CE 2 & related docs: SVC ESS improvement

52.2.1.1.1 JVT-W030 ( Prop 2.2.1) [X. Wang, J. Ridge] CE2: Improvement of MB mode pred in ESS

This proposal is a CE report on JVT-V108 with more results provided. In the current JSVM, inter-layer prediction on macroblock mode in ESS is based on partition information derived from base layer. More exactly, only if two blocks in an enhancement layer macroblock share the same partition from base layer, these two blocks can be merged into one. Such a method is asserted to tend to unnecessarily create smaller macroblock partitions and sub-partitions, which would in turn reportedly incur more interpolation complexity in motion compensation. JVT-V108 proposes a method in which two blocks may be merged into one as long as they share the same reference frame index and have similar motion vectors from the base layer. Further results provided in the report assert that the proposed method can effectively solve the alleged problem with essentially the same coding efficiency.

Try to combine base layer blocks into larger partitions for mode prediction (when reference index is the same and MVs are close in value). No change to coding efficiency reported. Significant reduction in use of small block sizes.

Remark: A “merge” has no effect if the motion vectors are equal and the block sizes are 8x8 or larger, so no text change is needed for that case.

JVT decision: Adopted (even where it may make no difference).

52.2.1.1.2 JVT-W058 ( Info) [E. Francois] CE2: Cross-check of JVT-W030 on ESS mode pred improvement

This document reports cross-check results of proposal JVT-W030 entitled ‘CE2 report: Improvement of macroblock mode prediction in ESS’ from Nokia. As a verification task, textual specification and corresponding JSVM software implementation were reported to have been verified and coding and decoding performance check was reported to have been carried out. The results presented in JVT-W030 were reported to be confirmed and the implementation within the JSVM software wass confirmed to match with the proposed textual specification.

Text was checked against software, software was available last time, software was tested. No problem reported.

52.2.1.1.3 JVT-W117 ( Prop 2.2/3.1) [Y. Ye, Y. Bao] CE2: Improved resid upsampling for ESS

JVT-V115 proposed a change to the residual upsampling process in ESS. In JVT-V115, the residual upsampling scheme makes the decision about whether to use bilinear interpolation or nearest neighbor copying based on the relative block alignment between the base layer and enhancement layer transform blocks. The scheme proposed in JVT-V115 was reported to (slightly) improve coding performance for commonly used ESS scaling ratios, and to improve visual quality. The contribution proposes a modified scheme that is mostly based on JVT-V115, along with another decision making based on the base layer block type (intended to further reduce blocking artifacts). The proposed scheme was reported to achieve small but consistent coding performance improvement over the reference JSVM_7_13 for all testing conditions

153

specified in CE2. Reconstructed video quality was reported examined and reported to also show visible improvement.

Changes when base layer residual edge (either 4x4 or 8x8) without an 8x8 enhancement layer edge – to use bilinear rather than nearest-neighbor – except when the edge is intra/inter.

Remark: Predictor (e.g., motion vector) may be different across that edge – predicting the residual across that edge seems questionable.

Addresses prior question of bilinear all the time negatively.

Basically no difference in PSNR measure quality – perceptual argument.

Significant visual improvement (subjectively) reported.

“Cherry picking” of results to report? Perhaps some.

See notes in section on JVT-W105.

52.2.1.1.4 JVT-W106 -QV (Late Info) [X. Wang] CE2: Verif Qualcomm JVT-W117 improved resid upsamp for ESS

The purpose of this document is to verify results in JVT-W117 from Qualcomm.

Compiled and compared PSNR results. Did not check for subjective improvement.

Question re: “Submission of final software and results [to CE partners]: next meeting – 2 weeks” – was this followed?

Proposal changed somewhat since last meeting – due to some artifacts discovered relating to intra/inter switch boundary.

See notes in section on JVT-W105.

52.2.1.1.5 JVT-W105 ( Prop 2.0) [X. Wang, J. Ridge] Study on residual upsampling without block boundary check under ESS

This proposal provides study results on the topic raised in JVT-V115. In JVT-V115, a method was proposed so that under ESS bilinear interpolation is performed across a base layer block edge if the edge falls within an enhancement layer transform block. By doing so two things may reportedly be achieved: 1) visual improvement (picture less blocky); 2) slight coding gain (about 1%). The contribution asserts that the for the case of ESS, doing bilinear interpolation without block boundary check can achieve essentially the same results claimed in JVT-V115. Visual quality is asserted to not show a visible difference from JVT-V115.

Residual prediction concept fails when predictor is substantially different across a base layer block edge.

Suggests that an encoder can detect situations where such a failure to create an adequate predictor may occur (e.g., MV discontinuity). In the reported test, intra macroblocks were conceptually assigned a zero motion vector value.

154

Remark: Testing for block boundaries, and making the upsampling process depend on that, seems to create a decoder burden. On the other hand, sometimes the “all the time” technique will require extra lower-layer residual blocks.

Remark: What about dyadic case? Response: The proposal considers only ESS. Remark: Treating dyadic as a special case seems undesirable from a design perspective.

Question: How about intra/inter switch boundary?

Remark: How about the Crew sequence? Response: Haven’t checked – the problem report had focused on Foreman.

Remark: The dyadic case should not be changed – there has been a lot of experience with that, and it creates more cases where extra residual block reconstruction can be avoided.

Remark: This (particular) encoder design does not avoid all artifacts – some failure cases remain.

Remark: It’s not clear whether the JVT-W117 method will avoid all artifacts either.

Remark: Bad failure cases should not be very difficult for an encoder to detect.

Suggestion: Failure cases are likely to remain, no matter what. Encoders will ideally need some kind of detection and avoidance. Primarily consider two factors: Decoder implementation friendliness, and stability and consistency of design.

Neither proposal, as proposed, changes the dyadic case.

Three main options considered:– Do nothing– As proposed in JVT-W117– As proposed in JVT-W105Opinions expressed were evenly divided between the three – no consensus for change.

No action taken on decoder text.

JVT decision: Adopt non-normative JVT-W106 encoder problem detection trick into JSVM.

52.2.1.1.6 JVT-W109 -LV (Late Info) [E. Francois] Verif JVT-W105 on residual upsampling without block boundary check under ESS

This document reports cross-check results of proposals JVT-W105 and JVT-W123 that both relate to residual upsampling in ESS. Both contributions propose solutions for reducing visual artifacts caused by residual upsampling. As a verification task, a coding and decoding performance check was reportedly carried out. The results presented in JVT-W105 and JVT-W123 were reported to have been confirmed.

The software implementing the proposals had been provided. Binaries were reportedly regenerated from these versions and used for generating the cross-check results, both for the original version and the modified ones.

The verification reportedly consisted of encoding and decoding, and checking that the provided figures of JVT-W105 and JVT-W123 fit with the results obtained.

155

Results data have been verified for the following configurations considered in document JVT-W105 and JVT-W123:– ratio 3/2: verification on bus, mobile, foreman and football sequences.– ratio 4/3: verification on crew and soccer sequences.– 3 layers: verification on crew sequence.

For all the performed verification tests, the encoder and decoder were reported to match perfectly. The decoded results were also reported to perfectly match the results provided in JVT-W105 and JVT-W123.

See notes in section on JVT-W105.

52.2.1.1.7 JVT-W123 ( Prop NN2.2.1) [X. Wang, J. Ridge] Analysis of visual artifacts in ESS residual pred

This contribution is a non-normative proposal that aims to address the issue of possible visual artifacts in ESS reported in JVT-V115. Detailed analysis of those areas with artifacts is asserted to reveal that the artifacts were caused by residual prediction with non-matching residuals. In this proposal, during encoding process such areas are identified so that a different R-D measure may be applied to prevent those visual artifacts. Results are asserted to show that with such a method the visual artifacts can be prevented while coding efficiency is preserved.

See notes on JVT-W105 and JVT-W117.

52.3 CE 3 & related docs: SVC subband coding

52.3.1.1.1 JVT-W097 ( Prop 2.2/3.1) [S.-T. Hsiang] CE3: Intra-frame dyadic spatial SVC based on subband/wavelet filter banks framework

This contribution reports CE3 results based on the previous contributions U133 and V084 that attempt to integrate the subband coding framework with the current JSVM for improved dyadic spatial scalable coding. Further simulation results are provided for Intra-coding under the CAVLC entropy coding mode. It also reports the results for dyadic spatial scalable coding under the long delay test condition utilizing the proposed algorithm for coding Intra frames only.

For intra-only coding average bitrate saving around 8% for QCIF-CIF, for uniform subband quantization in 4CIF 20% as compared to JSVM. For inter coding with long delay (where only I frame is wavelet coded) small loss as compared to JSVM.

Questionable whether for the intra case (where the lower layer uses a different reference) PSNR comparison is valid.

Only works for intra, dyadic, progressive. Not clear if it can be combined with bit-depth scalability.

This will not have a home in the current development. Very questionable where this would go in any future profile. Useful only for intra-only case, which is already covered in profile B intra.

JVT decision: Adopt to JSVM, but not with the automatic assumption that this will go into a draft by next meeting. Further evidence requested what it is good for; otherwise it may be removed.

156

52.3.1.1.2 JVT-W122 -QV (Late Info) [J. Ridge] CE3: Verif JVT-W097 wavelet-based intra dyadic spatial SVC

The results presented in JVT-W097 for the intra-only case were reportedly verified through compilation of source code provided by the proponent. The wavelet coder results reportedly appeared to match precisely.

There was reportedly a slight difference in the reference results, which reportedly appears to have been due to a difference in the JSVM software version used by the proponent and the verifier. However, this small discrepancy does not reportedly appear to have materially affected the conclusions.

Results for the long-delay case were not fully verified, reportedly due to time constraints.

Visual results at the highest layer reportedly correlate with the PSNR results. At the lower layer, the wavelet results reportedly are naturally sharper due to the difference in filter.

52.4 CE 4 & related docs: SVC bit-depth scalability

See also the closely-related ad hoc group report JVT-W010.

52.4.1.1.1 JVT-W102 ( Prop 2.2/3.1) [Y. Gao, Y. Wu] CE4: SVC bit-depth scalability simulation results

This contribution presents simulation results of bit depth scalability with the technique proposed in JVT-V061. This technical solution to bit depth scalability is asserted to be compliant to current SVC standard. The contribution indicates that there is no new syntax element needed to support bit depth scalability. Only a process of inter-layer bit depth prediction using fixed a left shift is invoked during the decoding process. The software integration and the test conditions in the performance test are subject to JVT-V304 and the conclusions from the Ad hoc Group of bit depth and chroma format scalability. Simulations were reportedly performed with eleven video sequences that covered a variety of bit depth/tone mapping approaches to create the 8-bit and/or 10-bit version from the same source video content. Detailed experimental results were also provided.

Remark: No actual energy in some of the higher bit depth video.

Remark: Seems like the most obvious way to do bit depth scalability. Related remark: This requires 10 bit motion comp in the enhancement layer (although still single-loop).

Remark: There was a competing proposal JVT-V078 at last meeting. That proposal used 8 bit motion comp and had multiple mappings to the enhancement layer, including the one in this contribution as one of them. It was planned to also be evaluated in the same core experiment, as reported in JVT-V304.

Remark: Test conditions and test material were available late, so the JVT-V078 proponents did not have adequate time for preparation of experiment results. JVT rules refer to a need for availability of necessary material by three weeks prior to the meeting and availability of final software and results by 2 weeks prior to the meeting. It was remarked that these deadlines were

157

not fully met, such that some material needed for the experiments was only available at the last minute before the JVT ordinary contribution deadline.

Question: Is the software part of the contribution? Response: It can be provided.

The contribution appears to show that the proposed method is an effective way of achieving bit depth scalability with a substantial advantage over simulcast.

We tentatively agree to accept the relative performance reported in this contribution, relative to single layer, as representative of the capability of the technique (in the absence of further evidence).

Continue CE – suggestion to crop pictures to 4CIF this time.

52.4.1.1.2 JVT-W116 ( Info) [A. Segall] CE4: Verif JVT-W102 (Thomson prop)This document reports the verification status of JVT-W102, which is titled "CE4: Bit-depth Scalability Simulation Results". The proponents provided Sharp with source code and simulations results. As of April 17, 2007, the verification is ongoing.

Experiments finished so far have successfully verified; more ongoing.

52.4.1.1.3 JVT-W113 ( Prop 2.2) [A. Segall, Y. Su] System for bit-depth scalable coding

A system for the scalable coding of higher bit-depth and/or larger dynamic range video sequences is reported. The approach is reportedly motivated by applications that do not utilize linear scaling to generate a lower bit-depth image from the higher bit-depth sequence. Examples include gamma correction, color correction, dynamic range limiting or other forms of tone mapping. The proposed design employs a modified inter-layer prediction scheme that consists of a series of shifts and adds (signaled in the bitstream like intra prediction modes) and addresses relationships between luma and chroma. The proposed process is spatially varying, and it is signaled in a manner similar to intra-prediction modes within AVC/SVC.

Suggests a reportedly efficient way of doing inverse tone mapping. Approach seems worth studying.

Intra-only results were shown. Benefit reported for inter-layer prediction mapping scheme. Asserts that either JVT-W102 approach or JVT-V078 approach can use this technique. Requests inclusion in CE.

It may be beneficial to test this concept in either scheme (JVT-W102 or JVT-V078).

52.4.1.1.4 JVT-W076 ( Prop 2.2) [J. Jia, H. K. Kim, H. C. Choi, J. J. Yoo] SVC chroma format scalability

From an investigation on the current SVC draft for chroma format scalability, it was reported that current SVC design works for most of the cases when the chroma format scalability is combined with the spatial, temporal or/and quality scalability – where both luma and chroma components in the enhancement layer are encoded. However, there is one case where the current standard draft was reported to not work well in terms of coding efficiency performance. That is when only chroma format scalability is applied to an enhancement layer. The current draft specification is

158

reported to have been designed to code all the information regarding the luma and chroma parts together for an enhancement layer while in a chroma-only scalability case, luma related information would not be required because that information is already coded in the lower layer.

Question: How much bit rate is saved by customizing for this case? Not reported. Group suggestion: provide results. Contribution noted.

52.5 SVC high-level syntax

52.5.1.1.1 JVT-W020 ( Prop 2.0) [Z. G. Li, S. Rahardja, S. L. Xie and W. Yao] Hypothetical reference decoder for video coding

This document provides two different methods for the conformance of a coded scalable bitstream to hypothetical reference decoder (HRD).

When digital video is compressed, the coded bit rate may vary significantly over time. The bitstream is sometimes transmitted over a reliable channel at a constant bit rate (CBR). While there is no packet loss in such a scenario, some jitter may occur among the packets. The buffer size of encoder picture buffer (EPB) associated with an encoding process and that of coded picture buffer (CPB) associated with the corresponding decoding process are finite. Hence the encoder must constrain the bit-rate variation such that a hypothetical reference decoder (HRD) with a predefined buffer size can decode the bitstream without resulting in any overflow or underflow (in non-low-delay operation). In the classical constant-delay mode, the coded data can be removed at the computed removal time while the decoding and display times preserve the output (possibly fixed) frame rate.

This contribution addresses the constant delay mode for both non-scalable video coding and scalable video coding (SVC). It asserts that the sending rate can be greater than the coding rate and there may be jitter, and the dynamics of EPB and CPB can be nonlinear because of the possible saturation, and the EPB and CPB may not be complementary. Iterative algorithms are designed for the HRD by taking both the jitter and the total size of coded bitstream into consideration. This is reportedly necessary to minimize the values of buffer size and initial buffer delay when there is saturation on the dynamics of EPB and CPB. An interpolation algorithm is also presented such that the coding rate and the sending rate are decoupled as in the prior design. SVC is composed of a base layer and (possibly several) enhancement layers, and each enhancement layer has its “base layer”. The conformance of each layer is proposed to be checked by defining the corresponding constraint for each layer by the proposed method. The base layer has two transmitted values: the buffer size and the delay between storing a picture in the buffer and starting the decoding of that picture. Each enhancement layer is proposed to transmit two values: the difference between the buffer size in the layer and its “base layer" and the difference between the delay in the layer and its “base layer" (using a coded difference to enable efficient representation). Two different methods are proposed for the HRD of SVC. In the first method, the sizes of all frames from the base layer to the current enhancement layer are used to compute the buffer size and the delay in the layer. In the second method, only the sizes of the current enhancement layer data are used. It is reportedly observed that the values obtained by the first method are usually smaller while those by the second one are more scalable.

Constant delay mode in case of channel jitter. EPB and CPB may no longer be symmetric then. Proposes two algorithms considering initial buffer fullness. For base layer buffer size and delay are transmitted, for enhancement layers differences in buffer size and delay as compared to corresponding base layers.

159

Only focuses on CBR case. Claim that main idea can still be used for VBR. In principle no problem with the current approach. General opinion that the amount of bits potentially saved by differential coding is not worthwhile to consider this.

HRD design is critical to the standard.

Focuses on CBR case.

Remark: What about the notion of multiple schedules?

Remark: SVC is designed as a single-loop syntax with data partitioning. Current syntax and HRD design conceptually can apply – refer to meeting report of last meeting (“submitted as JVT-V068r4. Other parts of JVT-V068 adopted (separate HRD parameters for each, and include temporal level in scalable nesting SEI).”)

Remark: Per meeting report of last meeting, current design seems conceptually similar to first proposed variant. Proposal suggests to code differences rather than totals, saving some bits at SPS level.

Remark: Differences are relative to what (considering multiple schedules)?

Remark: Amount of bits this would save is not a problem of a magnitude worth fixing.

Contribution noted.

52.5.1.1.2 JVT-W046 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Support for SVC header rewriting to AVC

It is asserted that there are two structures for coded video sequences that would allow lightweight SVC-to-AVC rewriting by removal of certain NAL units and NAL unit header SVC extensions as well as conversion of SVC VCL NAL unit types to the corresponding AVC ones. In the first structure, a temporal enhancement is provided as an enhancement layer to a Baseline profile base layer. In the second structure, more than one AVC stream is encapsulated within an SVC stream.

This contribution first proposes a change to the sequence parameter set SVC extension syntax and slice header syntax to enable the lightweight SVC-to-AVC rewriting. It is further proposed that syntax structure for the SVC-to-AVC conversion in the scalability information SEI message is appended with an indicator of the conversion operation, such that interoperability information of lightweight rewritten bitstreams can be signaled.

Proposal elements:1) Flag in SPS extension for “trivial” rewriting ability2) Alignment of SVC slice syntax with AVC slice syntax (seems OK, but let’s make sure)3) Scalability info SEI appended with “conversion type” info4) slice_type values, adding “all conceptually the same kind” indicationJVT decision: Adopted.

Question: Effect on deblocking filter of slice skip or other rewriting tricks?

Answer: If set base_mode_flag equal to 1 and residual_prediction_flag to 1 without sending coeffs, should inherit the CBP and QP and transform_size_8x8_flag from the base layer for deblocking purposes Also deblock IntraBL as Intra. Also follow this spirit if we notice similar

160

issues, conditioning on simple rewriting flag if in would be inappropriate not to. JVT decision: Agreed.

52.5.1.1.3 JVT-W047 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang] Pictures not for output in SVC

It is asserted in this contribution that there are two sources of needs for indicating whether decoded pictures are to be output. First, it is assumed to be a desirable feature that the layer with the highest dependency_id may be coded with a lower temporal resolution than its base layer. In such a coded stream, certain enhancement layer slices are coded as “skipped” and should not be output. Second, thinning of a scalable bitstream may result into a decoded sequence that is argued to be of insufficient quality for output in the presented coding schemes, logo insertion and discardable data adaptation. It is proposed that an output_flag is included in the SVC NAL unit header and controls whether the decoded picture is marked as “needed for output” or “not needed for output” in the decoded picture buffering process. It is additionally proposed that a syntax element layer_output_flag[ i ] is included in the scalability information SEI message to indicate which layers or operation points are not intended for output and hence should not be output.

JVT decision: Adopted.

52.5.1.1.4 JVT-W048 ( Prop 2.2.1/3.1) [M. M. Hannuksela, Y.-K. Wang, Y. Chen] On SVC high-level syntax

This contribution proposes 1) a change to the semantics of sub-sequence information SEI message to align with the latest definition of IDR picture, 2) a couple of constraints to the semantics of store_base_rep_flag and idr_flag, and 3) some syntax changes regarding presence of the syntax structure dec_ref_pic_marking_base( ).

1) adopted2) 2.1 adopted, 2.2 (removing constraint on base layer IDR needing enhancement IDR) adopted.3) Depends on use_base_representation, further studied during meeting, and adopted.JVT decision: Adopted.

52.5.1.1.5 JVT-W051 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] On SVC scalability information related SEI messages

This contribution first proposes two technical changes to scalability information SEI message among some editorial changes. The first technical change is inclusion of signaling for maximum number of buffered decoded frames and maximum number frames reordered for output for each scalable layer. This signaling enables a decoder to allocate minimum decoded picture buffer size for decoding a subset of the bitstream, and to start to output and display as soon as possible with the minimum initial delay. The second technical change is inclusion of signaling of profile, level, and bitrate information for quality layers. Furthermore, slight changes to some other SEI messages are proposed to enable using common SEI messages for both SVC and MVC.

JVT decision: Adopt signaling of profile and bit rate indication for quality layers. Do not adopt the “unification” with MVC (because it is far from clear at current point what MVC will need, and does not make sense to complicate SVC for that). For further issues, see under JVT-W064 below.

161

52.5.1.1.6 JVT-W052 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] SVC feedback based coding

It is assumed in this contribution that an encoder could avoid the use of the base representations for inter prediction, if it has received feedback from the far-end decoder indicating whether all the quality layers of the corresponding access unit were correctly decoded. It is asserted, however, that decoders have no means to conclude whether all the quality layers of a particular access unit have been received completely and decoded without mismatch. A quality layer integrity check SEI message is proposed for enabling the presented feedback-driven usage of the base representation in inter prediction. The message includes a cyclic redundancy check (CRC) code calculated over the NAL units for which quality_id is greater than 0. A change regarding the presence of the syntax element store_base_rep_flag is also proposed to enable the feedback based coding.

Remark: In case of MGS the decoder could potentially know the completeness when scalability info SEI messages completely received. In case of FGS it would not be possible. Alternatively, for MGS other more simple methods than CRC would be viable (e.g. signaling the maximum quality layer).

JVT decision: Adopted.

For a given quality ID, should macroblock data be required to be present for the entire picture? JVT decision: Only for quality ID = 0.

52.5.1.1.7 JVT-W137 -B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Revised syntax for quality layer SEI message

Reports modifications of quality layer SEI syntax needed for removal of FGS.

JVT decision: Adopt - Editors are also given discretion to “clean out” any further-identified remnants of removed features.

52.5.1.1.8 JVT-W053 ( Prop 2.0/3.1) [M. M. Hannuksela, Y.-K. Wang, D. Singer, T. Rathgen] SVC priority_id value setting method indication

SVC file format allows conveyance of multiple sets of alternative priority_id values for one SVC bitstream. A server may re-label the priority_id values for all the NAL units with one set of alternative priority_id values before sending the bitstream, thus to allow customized priority based adaptation. For each set of alternative priority_id values, a field priority_assignment_method_id is included to identify the method used to calculate the prioritiy_id values. This contribution proposes to include the indication of the priority calculation method for the default set of priority_id values contained in the NAL unit headers in the scalability information SEI message.

In principle useful. JVT decision: Adopt after revision of nt-string (per r2 of document).

52.5.1.1.9 JVT-W064 ( Prop 2.2/3.1) [J. Luo, L. Zhu, P. Yin, C. Gomila] VUI updates for SVC

This contribution proposes to modify the H.264/MPEG-4 AVC Video Usability Information (VUI) for the Scalable Video Coding (SVC) standard. The bitstream restriction information in VUI is independent for each interoperability point (IOP). This contribution aims at modifying the VUI to transmit bitstream restriction information for multiple IOPs. It is also considered how to

162

use SEI messages to convey bitstream restriction information for an H.264/MPEG-4 AVC compatible layer.

Related to JVT-W051. Difference is putting bitstream restriction in VUI vs SEI. After offline clarification with proponents of JVT-W051.

Seven bitstream restrictions proposed – identical to those already in VUI for entire bitstream. Proposed to specify them per-layer. Issue of how to handle base-layer temporal subsequences and SDP syntax.

Put in both places? No. Put them (all seven, with presence indicators as in current VUI) in scalability info SEI. JVT decision: Agreed.

52.5.1.1.10 JVT-W091 ( Prop 2.2/3.1) [L. Cieplinski] HRD parameters for SVC bitstream rewriting

An earlier contribution proposed the extending the Hypothetical Reference Decoder for SVC to include parameters to support bitstream rewriting for CGS. This contribution proposes an alternative way of incorporating the additional parameters, which is claimed to result in less significant changes to the specification.

Concern is raised about changing the picture timing SEI message. In principle, the HRD parameters could also be determined when re-writing is done and need not be transmitted beforehand. Application example / showcase about usefulness of the proposal needs to be provided.

Question: Is the rewriting process fully specified? Unless we can fully and clearly specify the rewriting process, how can we know what HRD parameters they will conform to? Don’t know which pictures the translator will choose to pass onward, which enhancement layers it will choose to include, etc. The same decoding process outcome may have multiple patterns of expression on AVC syntax.

Contribution noted. Intriguing, but unable to accept in this form – ideas would need more maturation.

52.5.1.1.11 JVT-W114 ( Prop 2.2) [A. Segall, J. Zhao] Showcase for transcoding scalability info SEI

In the Marrakech meeting, JVT-V036r1 was adopted to add AVC bit-rate information to the SVC Scalability Information SEI message. This contribution provides the required showcase of the SEI modifications.

Audience is satisfied with showcase.

52.5.1.1.12 JVT-W125 ( Prop 2.2) [G. J. Sullivan] On SVC high-level syntax and HRD

As SVC is designed as an extension of the AVC (ITU-T Rec. H.264 | ISO/IEC 14496-10) specification, it is important to consider the relationship between future SVC bitstreams and existing AVC decoders, and the relationship between different SVC decoders that are operating in the same system environment. It is also important to establish appropriate buffering and timing constraints to establish bitstream conformance, particularly including proper specification

163

of an SVC HRD. This contribution proposes several high-level syntax modifications and an HRD design to address these issues. As an additional “clean-up” remark, the contribution also suggests a modification to the definition of arbitrary slice order.

Issue 1 (SPS/PPS/SEI): Other proposals to address this were discussed during last meeting.

Remark: Similar reasoning applies to access unit delimiter NAL units – these could also be subsumed into an SVC NAL unit type (and assigned a D,T,Q).

Clarified offline with Miska and discussed further. Suggestion: Use prefix NAL unit to assign D,T,Q for SPS/PPS/SEI/filler (not AUD). Remark: Various implications discussed. No action.

Issue 2 (prefix NAL unit): Do not retain suffix NAL units. Type 14 prefix NAL units should always be used instead. JVT decision: Adopted.

Issue 3 (filler data): JVT decision: Adopted using prefix NAL unit to assign D,T,Q.

Issue 4: Make sure that the NAL header bytes cannot cause start code emulation. JVT decision: Adopted (exact form of header syntax to be determined).

Issue 5 (HRD): Possibility to add an informative clause about bitstream extraction (similar in spirit to figure under 3), but it must be guaranteed that the extracted subset is still a conforming bitstream.

Definition of VCL NAL units – should not have changed what an AVC non-scalable decoder will do with NAL unit type 20 the HRD. JVT decision: Agreed.

Issue 5 (ASO): JVT decision: Adopted.

Remark: Removal of SVC SEI messages rather than NAL units? No. Remove SEI NAL units not associated with any VCL NAL unit in the access unit - using prefix or content, which need to be consistent. JVT decision: Agreed.

Error resilience: 49, 50, 54, 62

52.5.1.1.13 JVT-W049 ( Prop 2.2.1/3.1) [C. He, H. Liu, H. Li, Y.-K. Wang, M. M. Hannuksela] Redundant pictures in SVC

Redundant picture support is one of the error resilient tools in H.264/AVC for enhancing the robustness to packet loss. Currently it is open whether to support the redundant picture feature for SVC enhancement layers. This document provides simulation results comparing different coding cases with or without coding of redundant pictures. It is proposed that the redundant picture feature is supported for SVC enhancement layers and included into the Scalable Baseline profile. Furthermore, an SEI message is proposed to contain redundant picture properties, based on which a decoder can determine whether the redundant picture can be used for inter-layer prediction when the corresponding primary picture is lost.

IDR in random access example can be realized by MGS enhancement picture?

Provides tests that seem to demonstrate usefulness of redundant pictures.164

JVT decision: Adopt SEI message contingent on adoption of redundant coded pictures in a profile.

Profile aspect open. Proposed to add to scalable baseline profile enhancement layers. JVT decision: Adopted.

52.5.1.1.14 JVT-W050 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela] On tl0_pic_idx in SVC

Document JVT-V140 proposed to remove tl0_pic_idx from the NAL unit header and include it in the RTP payload instead. In Marrakech, JVT decided to put the syntax element tl0_pic_idx at the start of the slice header and the prefix/suffix NAL units, and planned to adopt the JVT-V140 approach in San Jose if IETF AVT would take action to adopt the JVT-V140 approach to the SVC RTP payload format. In the March 2007 IETF AVT meeting, AVT adopted the signaling of tl0_pic_idx, among others, in the payload content scalability information (PACSI) NAL unit that can be present in the beginning of RTP packets. PACSI NAL unit is described in Section 6.10 in the latest Internet-Draft of SVC RTP payload format, available from http://www.ietf.org/internet-drafts/draft-ietf-avt-rtp-svc-01.txt. It is asserted that the latest Internet-Draft of SVC RTP payload format effectively satisfies the condition for adoption of JVT-V140 approach to SVC. Therefore, it is proposed that, in the SVC specification, tl0_pic_idx is not included in the NAL unit header or slice header and is only signaled in the SEI message as presented in JVT-V140 and copied in this document with minor editorial changes.

See notes below in section on JVT-W062.

52.5.1.1.15 JVT-W062 ( Prop 2.2/3.1) [A. Eleftheriadis, S. Cipolli, J. Lennox] Improved error resilience using temporal level 0 picture index

This contribution reviews the status and proposes further improvements to the concept of tl0_pic_idx. The field was introduced in the NAL unit header extension of SVC in JD8 (Hangzhou) to address the behavior of an SVC decoder (and SVC systems in general) in the presence of packet errors. It was shown that it is a way to use temporal scalability and multiple reference pictures to implement “zero-delay ARQ”, something that was not possible with earlier video coding systems. In JD9 (Marrakech), it was re-cast as an element of the slice layer (but in exactly the same bitstream location), in an attempt to have a fixed-length NAL unit header in SVC, with further action dependent on incorporation of the feature in the RTP payload format for SVC. The field was subsequently adopted into the RTP payload format for SVC in the March 2007 IETF meeting in Prague, together with two additional flags (that signal the first and last, respectively, NAL unit of a picture). This contribution first re-introduces the proposal for adding these two associated flags that signal the first and last, respectively, NAL unit of a picture, in order to address the case where the lowest temporal level picture data is transported over multiple NALs. It is shown that, coupled with RTP sequence number tracking, this design allows immediate detection of lost data for the lowest temporal level pictures both when no picture data is received, as well as when partial data is received. This contribution further describes three syntax designs for the tl0_pic_idx itself: a fixed-length NAL header, a variable-length NAL header, and a design in which the tl0_pic_idx field is moved to a new SEI message. It is shown that the fact that SEI messages can only appear at the beginning of an access unit, renders the SEI solution ineffective if SVC NAL ordering is strictly followed. It is also shown that a further limitation is the fact that SEI messages do not carry DTQ information in their NAL headers, whereas the Scalable Nesting SEI message does not provide the needed T information. Finally, the contribution identifies a bug in the current JD9 with proposed changes document, in that the tl0_pic_idx is not shown as a payload to a suffix NAL unit, as adopted in the Marrakech meeting.

165

Proposal to put tl0_pic_idx in SEI message..

Offline discussion result is documented in section 3.3 of revision JVT-W062r3.

JVT decision: Adopt section 3.3 of JVT-W062r3.

52.5.1.1.16 JVT-W054 ( Info) [I. Radulovic, Y.-K. Wang, S. Wenger, A. Hallapuro, M. M. Hannuksela] Multiple description coding using AVC redundant pictures

Multiple description coding (MDC) reportedly offers a competitive solution for video transmission over lossy packet networks, with a graceful degradation of the reproduced quality as the loss rate increases. This paper describes how redundant pictures, an error resilience tool included in H.264/AVC, can reportedly be employed in conjunction with MDC, in a standard compliant manner. It is asserted that comparisons with state-of-the-art techniques show a superior performance of the scheme, both in terms of an average PSNR, and in the smoothness of the reconstructed video.

Document for information only.

52.5.1.1.17 JVT-W068 ( Prop 2.2/3.1) [C. Tu, S. Srinivasan, S. Regunathan, G. Sullivan] CE4: 4-tap MC interp for high-res SVC enh layers

This contribution proposed a 4-tap motion compensation interpolation filter for SVC enhancement layers.  It is proposed to shorten the SVC motion compensation interpolation filters from 6 taps to 4 taps in order to reduce computational complexity. The 4-tap filter can reportedly be implemented using 16-bit only arithmetic. Coding performance was demonstrated, which is reportedly comparable (around 0.03 dB better on average) to the current 6-tap filters, and reportedly outperforming the H.241 RCDO interpolation method for 4CIF sequences. For CIF sequences, although on average performance penalty was reportedly around 0.2 dB, it was comparable to the 6-tap filters on some sequences. It was proposed to adopt this 4-tap motion compensation interpolation filter for luma for SVC high resolution (for example, standard definition and higher, or 720p and higher) enhancement layers, and to use it as an optional interpolation filter for low resolution enhancement layers.

Actual numbers of operations (assuming given distribution of positions) not given. No results with SNR scalability where the difference between interpolation results of base and enhancement might be critical. For low resolution (CIF) average loss around 0.1..0.2, maximum loss 0.5 for Mobile. For higher resolution loss is almost negligible (on average, varying between -0.1 and +0.1 dB).Proponent recommends to make this switchable.

Showed some loss for low resolution video; approximately neutral for high resolution. Spatial scalability and temporal scalability using B and P hierarchies.

Remark: How about SNR scalability? Not tested in contribution.

Remark: Comparative complexity analysis? Complexity of proposed filter is described in contribution, but not alongside a comparative analysis relative to the current MC interpolation method.

166

Remark: Test set seems limited.

Remark: Not much experience with 4 tap, would not be comfortable with complete replacement of current method.

Remark: Rewriting feature impact?

It only seems reasonable, considering the above, to consider adding as an additional supported feature rather than as a replacement.

Key pictures would need two motion compensations.

Not adopted.

Additional information uploaded in revision: Reported a 40% computation reduction and 23% memory bandwidth reduction for 8x8 block size.

52.5.1.1.18 JVT-W072 ( Info) [H. Schwarz] Results comparing JSVM, 4-tap, and RCDO MC interp.

In this contribution, simulation results comparing the coding efficiency of different sub-sample interpolation filters for the luma component in SVC enhancement layers were reported. The following interpolation filters had reportedly been tested: H.264/AVC interpolation filter as currently specified in the JD, RCDO luma interpolation filter as specified in H.241, and the 4-tap interpolation filter as proposed in JVT-V090.

Similar results as JVT-W068, but only reported for low-resolution case. Identical results to JVT-W068 for the overlapping test cases.

SNR scalability – QCIF and CIF.

52.5.1.1.19 JVT-W027 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of 4 tap motion compensation interp

This document is an information contribution on the evaluation of the 4 tap motion compensation interpolation proposed for SVC. The obtained results reportedly show that the 4 tap filter gives equivalent results to the 6 tap one on many sequences but may have some important impact on some specific sequences (Mobile QCIF/CIF with +12% rate increase, Bus QCIF/CIF with +4% rate increase), which visually corresponds to less sharp pictures. Consequently the contributors suggest to keep the 6 tap MC filter at least for SVC Profile B and to possibly consider the use of the 4 tap MC filter for SVC profile A, rather dedicated to mobile applications and more concerned by complexity issues.

This (information doc) reports similar results in terms of loss for high resolution (4CIF). Highest PSNR decrease in case of City. Thomson recommends to keep the 6-tap filter (or make it switchable) for profile B.

Bitstream rewriting would no longer be supported unless the 6 tap filter was used in the enhancement layer.

In case of use_base_rep flag, it would also need to be disallowed (otherwise would need 2 MC operations with the different filters).

167

Switchability would lead to more complex (in terms of gates) hardware. However, might save battery lifetime (how much?)

Similar results reported as in JVT-W068 and JVT-W072.

52.5.1.1.20 Discussion of potential rearrangement of NAL unit orderA top-down ordering of SVC NAL units was suggested and discussed. One mentioned issue relating to it was that an encoder would need to add delay for rearrangement of its bottom-up-generated NAL units into a top-down order. A decoder that receives things in a non-preferred order within an access unit could, if it wishes, operate by buffering up the access unit to achieve the processing order that its designer desires to follow. No action.

52.6 SVC applications and profiles

JVT decision: Branch out the software parts relevant for the current standard phase 1 as WD 1 of reference software. The remaining part of JSVM (with other tools) will be further maintained after that first step is done.

JVT decision: Editors are given discretion to put in any definition that we forgot to talk about.

52.6.1.1.1 JVT-W075 ( Prop 2.0/3.1) [M. Horowitz, A. Eleftheriadis] Max frame size for enh layers of SVC profiles <withdrawn>

This contribution presents a mechanism for modifying the constraint imposed by the maximum frame size (MaxFS) in the H.264 | AVC level specifications to support applications that require a large range of frame size and frame rate combinations at a particular level (e.g., video surveillance and video conferencing). For example, for a video surveillance application compliant with a given level it may be advantageous to decode 720x480 video frames at 15 Hz and at that same level decode 1920x1080 frames at 2 Hz. For this example, the MaxFS constraint requires that level 4 (or higher) be specified to decode the larger frame size where level 2.2 would have sufficed for both video streams without the constraint. In the proposal, the MaxFS column of Table A-1 is replaced with an expression for deriving MaxFS. The values for MaxFS are derived so that the resource requirements (e.g., MB/s, DPB, etc.) are level-for-level identical to the existing H.264 | AVC level structure. The context of this proposal is the enhancement layers of the SVC profiles. That is, it is neither being proposed for the SVC base layer nor for other existing profiles of H.264 | AVC.

Question: Current content of A.3.1 (similar in A.3.2).

e) PicWidthInMbs * FrameHeightInMbs <= MaxFS, where MaxFS is specified in Error:Reference source not found

f) PicWidthInMbs <= Sqrt( MaxFS * 8 )

g) FrameHeightInMbs <= Sqrt( MaxFS * 8 )

For example, Level 3 supports 5 pictures at 4CIF resolution. This proposal would also require support of one picture with five times that number of macroblocks – a “20CIF” picture, but at a five times lower maximum frame rate.

H.241, RFC 3984, and 3GPP documents have something related this (although not exactly the same).

168

Remark: But how can we do this for enhancement layers while keeping the base layer constrained by the original spec?

Enhancement layer might have a lower “level” than the base layer?

After further consideration, proposal withdrawn.

52.6.1.1.2 JVT-W093 ( Prop 2.2.1/3.1) [H. Chung, M. Karczewicz, J. Ridge, X. Wang, W. Han, S. Kim] SVC FGS profile

This document provides additional results to compare FGS with MGS in the so-called Profile C (not an actual currently-planned profile, but a further study topic) for SVC. It is claimed that AR-FGS offers the ability to respond to forced bit rate adaptation in a more graceful manner than MGS in a low-delay environment. Creation of a profile including FGS scalability is proposed.

Remark: Considering timing of work schedule and recent modification of MGS to improve its granularity properties, further analysis of FGS requirements should be postponed until the next meeting.

Remark: Current software does not support slice-structured coding. It may be difficult to maintain the current FGS and AR-FGS functionality in the software while working on proper support of “phase 1” features. It was suggested to branch the software and allow FGS and AR-FGS and other non-“phase 1” features to be removed or to cease to function properly in the “phase 1” branch. JVT decision: Agreed.

Contribution focuses on small frame variations due to characteristics of next-gen networks – reason: low-delay applications with some types of (e.g., CBR) characteristics. Particular interest was expressed by the contributor in AR-FGS. Comparative data shown reporting AR-FGS advantage in some cases.

Assertion is that our “phase 1” approach cannot be used to achieve nearly-constant frame size with low delay. Hierarchical P picture approach is asserted to be inappropriate due to variation in frame sizes.

End-to-end delay analysis? How much can delay be reduced and how much will remain? Proponent estimates 200 ms end-to-end delay.

Bit rates? Frame rates?

Potential for feedback usage.

Potential for taking enhancement picture into account for key pictures.

Appropriate content? (Is the Bus sequence really relevant?)

Set up an AHG on identification of application requirements for FGS and simplification of FGS design.

52.6.1.1.3 Profiles definition changes

169

SVC Profiles tools tableScalable Baseline

a.k.a.

SVC A

Scalable High

a.k.a.

SVC B

Scalable High Intraa.k.a.

SVC B Intra

AVC base layer (dependency_id equal to 0 and quality_level equal to 0) Profile

Baseline High High

Impacting AVC base layer tools slice_type I, P I, P, B Ideblocking filter Y Y n/aconstrained_intra_pred_flag in base layer

1 1 1

num_slice_groups > 1 N N Nslice_group_map_type N n/a n/adirect_spatial_mv_pred_flag n/a 1 n/aarbitrary slice order N N Nredundant slices N N N

SVC tools slices I, P, EI, EP I, P, B, EI, EP, EB

I, EI

smoothed ref inter pred N Y YPR slice motion refinement N N NAR-PR slices N N Nfgs_coding_mode N N Ninterlace N Y YCAVLC Y Y YCABAC Y* Y Ydeblocking filter Y Y n/adeblocking filter (upsampling) Y Y Yconstrained_intra_pred_flagbelow the top layer

1 1 1

arbitrary slice order (within slice group)

N N N

num_slice_groups > 1 Yslice_group_map_type 2 N Nresolution factors 2, 1.5 Y Y YESS (any factor) N Y YESS aligned crop window Y Y YESS non-aligned crop window N Y YEIDR Y Y n/aIROI N N Nfragmented PR slice N N NCGS with varying quality levels (MGS)

Y Y Y

weighted prediction Y Y Yuse_base_representation_flag Y Y Y8x8 transform block size Y* Y Yquant scaling matrices Y Y Ynum temporal levels 8 8 8num dependency id 8 8 8max num decoded dependency id (using inter-layer prediction)

3 3 3

num quality levels 16 16 16color_bit_depth, color format 4:2:0/8 4:2:0/8 4:2:0/8

*: activation of the CABAC and 8x8 transform block size tool is subjected to levels definition (Level 2.1 (2CIF) and above)

Max NAL unit size (NumBytesInNALunit)? No.Smoothed reference prediction (see JVT-W026, JVT-W118, JVT-W126).B pictures in scalable baseline enhancement layers? MinLumaBiPredSize? See below.

170

For both SVC A and B, when PicSizeInMbs is greater than 1620, the number of macroblocks in any coded slice shall not exceed MaxFS / 4, where MaxFS is specified in Table A-1 (or SVC equivalent). JVT decision: Agreed.

For both SVC A and B, cpbBrVclFactor = 1250 and cpbBrNalFactor = 1500. JVT decision: Agreed.

Scalable High: Same level limits as High. JVT decision: Agreed.

Scalable Baseline:– Levels 2.1 and 2.2 SliceRate = 22 (which slices count? the slices for the layers that are

“necessary” for decoding, as can be determined from high-level syntax)– Allow B pictures (direct_8x8_inference_flag = 1 always, MinLumaBiPredSize = 8x8

always).– Define MaxSubMbRecSize 576 up to level 3, 1152 level 3.1 and 3.2, 1440 levels 4 to 4.2, no

limit for level 5 and 5.1 (limit for base layer too, and also enhancement layer).JVT decision: Agreed.

52.7 SVC other normative design proposals

52.7.1 SVC restrictions on interlaced coding

52.7.1.1.1 JVT-W025 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Restrictions on interlaced coding in SVC

This document relates to SVC interlaced video coding. The initial design of interlaced coding in SVC permits any field / frame picture configuration for the different layers. This proposal aims at introducing some constraints on these possible set of configurations, in order to ease the implementation of interlaced coding in SVC.

Three elements:1) Suggests to force field_pic_flag and bottom_field_flag to be equal across layers (while

allowing frame_mbs_only_flag to be different across layers).2) Suggests that base_frame_and_bottom_field_coincided_flag and

base_bottom_field_coincided_flag to be identical or to be removed (enforcing alignment of the top of the frame with the top field of the other layer).

3) Correct a clear error in some position calc equations.

JVT decision: Adopted.

52.7.2 SVC smoothed reference prediction

52.7.2.1.1 JVT-W026 ( Prop 2.0/3.1) [E. Francois, V. Bottreau, J. Vieron] Profile SVC B: Evaluation of smoothed ref pred

This document relates to the evaluation of smoothed reference prediction (SRP) on a number of various configurations, addressing both low delay and long delay coding applications. SRP is currently only considered for profile SVC B, mainly oriented toward broadcast applications (it is not included in profile SVC A that rather relates to mobile and real-time applications). The obtained results reportedly show a slight bit rate increase by removing SRP on most sequences

171

(for long delay configurations: average 0.70%, worst case 2.32%; for low delay configurations: average 1.94%, worst case 5.99%), not noticeable on a visual quality point-of-view. The less favorable results for SRP are observed for long delay configurations that rather correspond to profile SVC B applications. Consequently the contributors recommend removing this tool from profile SVC B.

Reports that there is some PSNR benefit, mostly at high bit rates, but that perceptually the SRP has some undesirable excess smoothing effect.

Question: Were the frames selected to show perceptual issue “cherry picked”. Response: Basically, yes – as far as typical behavior with motion video running at full speed, basically see no difference in quality – not asserting that there is typically any significant difference in quality that way. But when viewing individual pictures, author asserts that some loss of resolution is observed.

Dyadic two-layer hierarchical prediction for “low delay” (2) – why? Just didn’t have time to try other cases.

Proposes not to use SRP in Profile B (the only profile it is currently in).

Author says that the tool requires some complexity to support. Remark: There are comments about that in another contribution.

52.7.2.1.2 JVT-W118 ( Prop 2.2) [Y. Ye, Y. Bao, W. J. Han, S. Y. Kim] Perf and complexity of smoothed ref pred

In the current SVC baseline profile, smoothed reference prediction is not supported (assertedly due to concerns over its complexity and performance). Further experiments have been carried out within the coding framework of SVC baseline profile. It is asserted that smoothed reference prediction not only offers notable performance gain that effectively reduces the gap between multi-layer SVC coding and single-layer AVC coding, but also reduces system complexity, both in terms of reducing computational complexity and reducing memory bandwidth requirement. Furthermore, it is asserted that smoothed reference prediction provides much better visual quality in the reconstructed video. Additional implementation cost is asserted to be low compared to the benefits it offers. In addition to supporting smoothed reference prediction in the scalable high profile, it is proposed to also enable smoothed reference prediction in scalable baseline profile.

Test reported: Dyadic spatial scalability, CAVLC (asserted to be a pessimistic scenario for SRP), three-layer hierarchical P.

Best case was 0.5 dB+ for Harbor, worst is City with basically no difference in fidelity measured. Average roughly in 0.2 to 0.3 dB range – better for CIF to 4CIF than for QCIF to CIF.

Reported conclusion is that when SRP is available, BLSkip is used more often rather than using other prediction modes that are more complex than it is, to the extent that the overall complexity is reduced around 15-20% as a percentage of inter prediction generation for the luma component. About 7% memory bandwidth reduction also estimated.

Visual example shown where extra detail is enabled when SRP is present. Another visual example of “noisiness” when not using SRP.

SRP is, basically switched as a submode of BLSkip with residual prediction – syntax at the MB level when that case is encountered 9 (and can be disabled at the slice level).

172

(Bitstream rewriting flags are also at slice level.)

There were prior contributions at the two previous meetings saying similar things.

It is currently in one profile but not the other. At the moment we don’t have a strong consensus that the tool should be put everywhere or removed from the standard.

Question: Two issues with IPR statement in contribution:1) Contribution is a proposal for a technology that was previously reported with a 2.2 IPR statement, but proposal has a 2.0 IPR statement? Response: Probably the proposal should have had a 2.2 IPR statement – will revise.2) Contribution is a two-company proposal, but only one company is listed in the IPR statement? Response: Second company will be contacted to clarify.

Revision uploaded Tuesday with Qualcomm 2.2 and Samsung 2.2/3.1 statements.

Further results later presented, asserting that if the encoder biases its decision-making to favor the selection of smoothed reference prediction, complexity is further reduced with no apparent impact on coding efficiency.

Has the tool been tried with interlace? Has interlaced support been stable in S/W?

What if the decoder impact if SRP is turned on for all macroblocks? Impact complexity does not seem large.

Remark: Consider complexity of needing to support switching between two inter-prediction modes – Response: Consider that switching of MC interp process now needs to happen at a finer granularity to support smaller block sizes.

Overall coding efficiency benefit of SRP is small – concentrated at high bit rates. Visual excess blur sometimes reported at such high rates.

Visual benefits shown for particular still-frame cases – it did look good in those examples.

Encoder complexity increases – but not by much since applies to testing only one case (BLSkip with residual prediction).

More benefit seems to be in low delay scenarios.

In terms of implementation effort, optimization effort, testing effort, and quantity of text, SRP adds a burden. However, in terms of processing cycles and other such measures, it may be somewhat statistically beneficial.

Remark: SRP helps reduce gap between SVC and single layer.Remark: Consider multi-layer optimization – e.g., per JVT-W071.

Tested GOP size = 4. Why not others? Just testing effort.

Remark: Did not find gain with larger GOP size. Larger GOP sizes are expected.

Remark: Internal testing by another company has led to a negative opinion of the feature.

173

Remark: Adds a difference relative to Base design.

Remark: This feature is an extra implementation burden for supporting scalability. We are not designing the base layer here. Design consistency is desired, and implementers of AVC decoders should not be burdened with a need to implement extra features to support scalabilty.

Available data for assessing usefulness of feature is limited.

SRP is not in Profile A and it seems clear that there is no consensus to add it.

Upon further discussion, SRP should be removed from Profile B. JVT decision: Agreed.

SRP can be considered as a Phase II investigative tool.

52.7.2.1.3 JVT-W126 ( Info) [Z. He] Verif JVT-W118 perf and complexity of smoothed ref pred

This report is to verify the document JVT-V118 “Performance and complexity of smoothed reference prediction in SVC profile A” from Qualcomm. Verification was performed based on the source code and configuration provided by Qualcomm, and the simulation results were confirmed for all the eight CIF sequences. 4CIF was partially verified (verified for the cases that were tested).

The reason the source code was from Qualcomm was due to not knowing the exact status of JSVM work and wanting to control off/on for SRP – the software is available on the ftp site as part of JVT-W118.

52.7.2.1.4 JVT-W112 -L (Late Prop 2.2) [A. Segall] Clarification of base_mode_flag <withdrawn>

Contribution JVT-W112 was submitted late, but was withdrawn as moot after some discussion, in consideration of action taken in response to other contributions, as is also noted elsewhere in this report.

A change is requested for the case that the base_mode_flag is one and adaptive_prediction_flag is zero. The fix enables the smoothed reference prediction process when it is enabled in the baselayer.

Current syntax seems to allow “weird” multi-layer combinations of motion vectors and the associated interpolation process with respect to smoothed reference prediction.

Proposes to infer the smoothed reference flag from the base layer when base_mode_flag = 1 and adaptive_residual_prediction_flag = 0.

Discussed offline. Proposal withdrawn, considering removal of SRP from Phase I.

52.7.3 SVC deblocking

52.7.3.1.1 JVT-W061 ( Prop 2.2/3.1) [D. Hong, A. Eleftheriadis, O. Shapiro] Modified deblocking filter process in scalable extension

174

This contribution introduces a modified deblocking filter process in scalable extension (subclause G.8.14 in JD9). The current process is mainly derived from the AVC deblocking filter process with modifications proposed by contributions JVT-O067 and JVT-P013. These previous contributions adjust the original AVC deblocking filter process to change the handling of the cases where base layer residue or sample values are used to derive current layer samples. The present contribution further adjusts the deblocking filter process by modifying the qPav derivation method so that the base layer QP is used for the deblocking when the enhancement layer blocks have no transform data, the residual of the blocks is predicted from the base layer, and (in the case of inter blocks) the enhancement layer blocks have similar motion vectors with the same ref_idx. This adjustment of the qPav derivation method was introduced in JVT-V089 where the arithmetic mean of the base and enhancement layer QPs was used, rather than just the enhancement layer QP. This contribution further considers the effect of applying the proposed qPav derivation method for various GOP sizes (2, 8, 32) for both hierarchical P and B structures. This contribution also tests and compares using several different types of weighted average combination of the base and enhancement layer QPs instead of taking just the simple arithmetic mean. Using just the base layer QP (an extreme case of the averaging where the base layer QP is weighted by 1 and the enhancement layer QP is weighted by 0), the modified qPav derivation method provides experimental results under the JVT common conditions that range from the maximum benefit of +0.654 dB PSNR to the maximum penalty of -0.003 dB.

Like JVT-V089, but using base layer QP instead of the average.

A significant PSNR benefit was reportedly shown when enhancement layer has much larger step size (QP increase by 15) than the base layer.

Remark: How about just turning down the deblocking filter strength? Reply: Can do that, but increases overall blockiness.

Fixed QP. Remark: Realistic?

Remark: Adding more conditions to deblocking filter. Response: Similar conditions to what the encoder is already using in the DF process.

Proponent asserts that this is in response to an issue that arose in an actual real-time implementation with rate control.

Question: How often does this issue arise.

Visual effect shown for a difference of 15 in QP. Data for a smaller QP difference requested.

No verification contributed. Text and software and bitstreams are (or soon will be) available.

Other experts were asked to study the proposed technique during the meeting.

Remark on somewhat related topic: What about RCDO deblocking? Has been put into software but not studied. Has not yet been shipped in products. Suggestion that some adjustments for SVC might not be appropriately made as-it-is.

Remark: Inheriting QP from base layer can help “rewriting” – suggest that using the base layer QP value when MB is not coded makes sense.

Remark: Experimented with it using QP+15 – helped in some areas and did not help in others. Also found that when the QP difference is large, upsampling the base layer can sometimes look

175

better than adding a very coarsely quantized enhancement layer. Overall impression was negative.

Idea: Send a “gamma” weight fraction sent at slice level like alpha and beta (in units of one eighth) to determine the weighting of QP between the base and enhancement layer: ( QP1 * f + QP2 * ( 8 - f) + 4)  >> 3. Default behavior is what is in current JD. Don’t send when doing rewriting (use enhancement layer QP in that case).

Try to find a method, such as the above, to adjust the effective QP that can capture the benefit of the current and proposed methods.

Remark: Goal of the weighting idea was to capture the benefits of each approach. Suggestion is for encoder to use the existing adjustment controls of deblocking filter process. This is asserted to suffice without a need for the further adjustability.

No action taken.

52.7.3.1.2 JVT-W063 ( Prop 2.0/3.1 Layered Media, then 2.2 from Polycom) [D. Hong. A. Eleftheriadis, O. Shapiro] Deblocking filter for SVC to support multi-threading with slice boundary

This contribution proposes to modify the current SVC deblocking filter process to support multi-threading, without having to turn off slice boundary deblocking. With the current process, a picture has to be sliced and deblocking across slice boundaries must be turned off in order to run deblocking of each slice in parallel. This creates an annoying “blockiness” artifact across slice boundaries in decoded images.

The desire is multi-threaded deblocking, which is difficult in the current design. Proposes to change the order of edge processing, so that right-to-left and top-to-bottom edge ordering is used.

Remark: May change access pattern of some hardware designs that are highly-customized to the current design.

Remark: Can switch the order of the interior edges again, reducing the number of stages further – from 4 to only 2 stages (horizontally and vertically).

Remark: Any perceptual effects? Proponent has seen no subjective difference – can provide sequences.

Remark: Effect on “bitstream rewriting” capability? Suggestion: Disable for rewriting-oriented coding.

Idea from Polycom: New value of deblocking_disable_idc that indicates applying filtering inside of slice first (without changing edge ordering), followed by filtering across the slice boundaries. Do not use this value when rewriting is enabled.

Revised contribution uploaded to reflect that (with a 2.2 patent statement from Polycom).

Remark: That’s OK, because it enables paralelizable encoding – focus at the moment is not on the decoding.

JVT decision: Adopt the idea from Polycom documented in revised (JVT-W063r1) contribution.

176

52.7.3.1.3 JVT-W069 ( Prop 2.2/3.1) [Z. He] Simplified H.264/AVC deblocking filter for SVC enh layer

This contribution proposes a reported simplification of the existing deblocking filtering for SVC enhancement layers while reportedly maintaining the same data and control flow as used for base layer. In reported results for four CIF@30fps common sequences, the proposed simplified algorithm reportedly shows a reduction on the data access and computation complexity by 60% in average compared to the original deblocking algorithm, with luma PSNR degradation about 0.03 dB (maximum 0.05 dB). In comparison, the RCDO deblocking reportedly has ~45% reduction with luma PSNR degradation of 0.1 dB (maximum 0.18 dB). Since the proposed deblocking reportedly has the same data- and control-flow as the existing H.264/AVC deblocking filter, the deblocking design can reportedly be shared in SVC base- and enhancement-layers.

Modifications:1) Only use BS = 0 or 12) Only use edge detection for one of the four rows3) Include an offset into the edge detector (specified in text or by encoder-sent syntax)

Very limited testing (only CIF).

There is not sufficient information available to make such substantial changes to the deblocking filter.

Contribution noted.

52.7.3.1.4 JVT-W128 -QV (Late Info) [Y. Ye] Verif of JVT-W069: Simplified deblocking for SVC enh layer

This document reports verification results for the proposal by Freescale as reported in document JVT-W069 “Simplified deblocking for SVC enhancement layer”. The results reported in JVT-W069 were reportedly confirmed.

The simulation results reported in JVT-W069 were reportedly verified. Out of the four sequences, Bus, Mobile and Foreman were reported fully verified; the results and the verification results reportedly matched exactly for these three sequences. For Football, the reported results and the verification results showed very small difference of up to 0.05 kbps, and the reason was reportedly probably due to different platforms being used in simulations (Unix vs. Windows XP).

52.7.4 SVC spatial scalability resampling

Ad hoc group finished its evaluation, and did not find evidence of a need for additional upsampling filters.

52.7.4.1.1 JVT-W028 ( Info) [E. Francois, V. Bottreau, J. Vieron] Evaluation of flexible 4-tap upsampling filters

This document is an information contribution on the evaluation of the adaptive upsampling filters proposed in JVT-V074. The results were asserted to show that on the tested sequences, the flexible upsampling filters do not provide significant improvements compared to the current non-adaptive solution.

Ran the software that supports alternative filter selection. Found no significant benefit.

177

Remark: The software does not include an encoder method for selecting which filter will be used, so no benefit would be anticipated from running the test that way.

52.7.4.1.2 JVT-W022 ( Prop 2.2/3.1) [T. Tran, L. Liu, P. Topiwala] Dyadic spatial down- and up-sampling filters for SVC

This proposal presents updated results for FIR low-pass filters that can be employed as dyadic down-sampling and up-sampling filters in SVC. The proposed filters reportedly have their roots from the wavelet and spline interpolation theory which is asserted to have long been established to have stable interpolation characteristics. All of the proposed filters have integer coefficients; some are asserted to have very low dynamic range and to be suitable for efficient VLSI implementation. This proposal also asserts that coding efficiency does not necessarily have to be sacrificed by employing short low-complexity integer-coefficient filters.

Proposes different downsampling and upsampling filters. Contribution supports both dyadic and ESS, but focuses on dyadic, since most benefit reportedly found there. Focuses on intra.

Results only provided in contribution for one sequence. Some other results shown that were not previously presented.

Proposal asserted to be the same as JVT-V030 / JVT-V031.

Performance asserted to be more measurable for low QP and intra-only. No gain for high-delay case. No significant gain for ESS cases – contribution focuses on dyadic.

Modified downsampler (odd-length mirror-symmetric). Proposed 4-tap phases are in JVT-V031.

Visual and PSNR benefits reported – esp. for base layer upsampled. Also some for enhancement layers – esp. at high bit rates.

Visual demo shown – base layer upsampled using provided filter when used with proposal’s downsampling filter used.

Base layer should perhaps not be watched as-is without such “matched” upsampling. Mixed opinions in favor of JSVM filter expressed in such experiments in Hangzhou.

Proposes to enable encoder selection of the proposed 4-tap upsampling table as an alternative to the current table.

Remark: Differences likely primarily due to downsampling change. Upsampling actual tap values very similar.

Remark: Position calculation for luma different than in reference.

Remark: Some phases (1, 5, and 7) were not tested in AHG activity.

Remark: Conceptually-reversed and significantly off phase positions in current design for linear ramp phase measure – specifically, one person from Microsoft expressed an opinion that phases 5, 6 and 7 from the prior Microsoft proposal or from this proposal looked better than the ones in our current draft.

Some question of application need for such upsampled base layer use.178

Some confusion over starting phase offset (lack of adjustability in current design for luma).

Phase alignment of luma is not adjustable in the current design. Adjustability would add requirement for all positions to be supported in the decoder – even when using fixed upsampling ratios like 3/2.

52.7.4.1.3 JVT-W086 ( Prop 2.2/3.1) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] Some consideration on the up-sampling position calculation

The up-sampling operation currently found in Joint Draft for extended spatial scalability (ESS) uses a particular method of calculating the position and phase information when up-sampling the low resolution layer. That method relies on the use of an approximate ratio computation. This contribution presents some implementation issues related to this approximate ratio. For some practical up-sampling ratios, non static phase pattern for up-sampling occurs which is not a desired feature for low complexity implementation. This contribution recommends then to specify use of an actual division operation in the position calculations for ESS rather than the use of the approximate ratio. Software and excel file reporting practical scenarios with identification of phase differences accompanies the contribution for illustration.

Group requested interested parties to confer off-line. Result was reported in JVT-W136.

52.7.4.1.4 JVT-W136 -B (BoG) [G. J. Sullivan, S. Pateux] BoG report on JVT-W086Summary of BoG overview conclusions regarding JVT-W086.

Presented.

JVT decision: Keep method as in current draft. For levels having picture width or height greater than 2048, specify scaling the resampling ratio up more by the constant amount that will still keep all calculations within 32 bits.

52.8 SVC non-normative contributions

52.8.1 SVC editorial input

52.8.1.1.1 JVT-W070 ( Text) [H. Schwarz, M. Wien] Editors input for SVC draftDraft text from the editors showing the current status of SVC text drafting work.

Shows progress in editing work – should be the basis for the future work. JVT decision: Agreed.

52.8.1.1.2 JVT-W099 ( Info) [J. H. Park, Y. H. Kim, B. H. Choi] Clarification of mb_qp_delta syntax

This contribution reports the clarification of mb_qp_delta in macroblock layer syntax in scalable extension to eliminate an unnecessary condition check. And also reports that no modification is needed on JSVM S/W.

179

Remark: Had a 2.2 IPR statement and was marked as a proposal, later revised as an information document with no attached IPR statement. Appears to be strictly editorial input.

Editors are asked to consider the comment in their drafting work.

52.8.2 SVC tutorial material

52.8.2.1.1 JVT-W132 -B (Requested Info) [T. Wiegand] Overview paper and presentation on SVC

This contribution, submitted at the request of the JVT, provides tutorial information on the SVC extension design for AVC.

52.8.3 SVC encoder and extractor optimization

52.8.3.1.1 JVT-W071 ( Info) [H. Schwarz, T. Wiegand] Further results for an rd-opt. multi-loop SVC enc.

The main disadvantage of the JSVM encoder control for multi-layer coding is that the losses against single-layer coding are unevenly distributed between base and enhancement layer. In JVT-T080 the basic idea of a joint multi-loop encoder control for spatial and SNR scalable coding has been described and first simulation results for IPPP have been shown. In this contribution, further results for hierarchical B pictures and a newer version of the JSVM software are provided. The simulation results demonstrate that enhancement layer coding efficiency can be traded-off for base layer coding efficiency.

For the cases of spatial and SNR scalability that were tested in this contribution, it was reported to be possible to adjust the coding efficiency for base and enhancement in a way that the rate increase relative to single-layer coding is about 10% for both the base and enhancement layer.

Remark: The last sentence above is approximately the same as saying that the goals of the SVC project have been fulfilled (in PSNR measure terms).

Basic idea is to jointly optimize the base and enhancement layer coding parameters (by an adjustable amount controlled by a weighting factor).

Shows how to measure the “usage” of the base layer rate in terms of its effect on the enhancement layer fidelity.

52.8.3.1.2 JVT-W029 ( Info 2.2.1/3.1) [W.-H. Peng] Low-complexity mode decision algorithm for combined CGS and temporal scalability

This contribution presents a layer-adaptive mode decision algorithm and a motion search scheme for scalable video coding (SVC) with combined coarse granular scalability (CGS) and temporal scalability. To speed up the encoder while minimizing the loss in coding efficiency, the “computational redundancy” between the coding layers is considered. Depending on the macroblock (MB) coding modes and the quantization parameters (QP) of the reference/base layer, a look-up table is recursively used to determine the MB modes to be tested at the enhancement layers. In addition, to avoid exhaustive motion estimation, the reference frame indices of the base layer are adaptively reused, and according to the MB partition at the

180

enhancement layer, the initial search point for motion estimation is selected from the motion vector at the base layer or the motion vector predictor at the enhancement layer. The proposed schemes were tested with standard sequences in CIF and 4CIF resolutions using 1 base layer, 3 CGS layers, 3 reference frames, and GOP size of 8 and 16. As compared with the mode decision algorithm in JSVM 8, the proposed schemes reportedly provide an average of 76% improvement in overall encoding time with an average increase of bit rate below 1%, and an average Y-PSNR loss below 0.01 dB.

Binary executable offered (not source code, at least not now).

Contribution describes a number of techniques that may be useful in fast encoder design. No action requested. Further investigation of such techniques, along with source code and verification, could potentially lead to a good low-complexity mode of JSVM software operation.

52.8.3.1.3 JVT-W043 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control for the Joint Scalable Video Model (JSVM)

The consideration of rate control algorithms within video encoding systems is very critical for a variety of applications where transmission may be constrained due to the channel’s bandwidth. Nevertheless, the authors have observed that all evaluation of the Joint Scalable Video Model (JSVM) reference software, and consequently of the Scalable Video Coding (SVC) standard, has been limited in experiments using fixed and pre-determined quantization parameters (QPs). Furthermore, very few, if any, experiments were performed to evaluate the impact of rate control to the scalability features of SVC. To this purpose, this contribution introduces the quadratic rate control scheme that has already been adopted within the H.264/MPEG-4 AVC Joint Model (JM) reference software in the latest JSVM software. This implementation only affects the SVC base layer but the scheme could be extended in the future to also support scalability layers as well. Although it could be arguable whether this algorithm can be considered as state of the art, the provided experimental results demonstrate that its Rate Distortion performance is equivalent compared to the use of fixed QPs, while achieving the target bit rate. This suggests that this tool should be a valuable addition within the JSVM software.

Software available – has been uploaded.

Question: Did they try the quality level assigner? No.

JVT decision: Adopted (integration with lower priority than normative things).

52.9 SVC conformance

52.9.1.1.1 JVT-W138 -B (BoG) [V. Bottreau] Toward an SVC conformance specification

Coordinators: Alex Eleftheriadis for Scalable Baseline, Vincent Bottreau for Scalable High and Scalable High Intra)

Every coding “tool” must have some conformance bitstream(s). Otherwise the tool will be removed from the specification.

Draft spec in manner similar to AVC conformance spec.

Common SVC features listed.Profile-dependent SVC features listed.

181

JVT decision: Plan approved per JVT-W138.

52.10 SVC verification testing

52.10.1.1.1 JVT-W110 ( Info) [E. Francois, V. Bottreau, J. Vieron] SVC verif test plan: Updated results for SVC High Profile intra

This is an information contribution that presents updated results according to the Draft SVC Verification Test Plan Version 2.2 (MPEG output document N8903) for supporting SVC Profile High Intra as defined during the last (Marrakech) JVT Meeting for Professional video manipulation scenarios.

52.10.1.1.2 JVT-W131 -B (Late Info) [D. Hong, A. Eleftheriadis] Verification bitstreams for SVC Profile A

This information contribution provides verification bitstreams for SVC Profile A, particularly for videoconferencing.

52.10.1.1.3 JVT-W135 -B (BoG) [I. Amonou, N. Cammas, S. Kervadec, S. Pateux] On SVC verif test plan

Summarizes conclusions from break-out. JVT decision: Plan approved. Refinement by the editors is invited.

53 Multi-view coding

53.1 CE 5 & related docs: MVC illumination compensation

53.1.1.1.1 JVT-W024 ( Prop 2.2/3.1) [W. S. Shim, M. W. Park, G. H. Park, D. Y. Suh, H. S. Song, Y. H. Moon, J. B. Choi] CE5 results- joint prop for MVC deblocking

In this contribution, CE5 results of the joint proposal of MVC deblocking for illumination compensation are reported. The joint MVC deblocking method (combined with JVT-V033 and JVT-V051) for diminishing or eliminating blocking artifacts caused by illumination compensation is asserted to be able to improve the subjective picture quality as well as maintaining the objective picture quality of the MVC video sequences.

Joint proposal from JVT-V033 and JVT-V051. Control bs by IC_flag and IC_offset to avoid additional blocking artifacts. Additional decision is included at the end of the bs=0 derivation. Bit saving average about 0.05%, mostly claim for subjective improvement. Most effects visible in flat areas (like Race). Subjective viewing (performed by Tobias last meeting) did not conclude for subjective improvement in cases of ballroom and exit (only 3 sequences were tested).

The group checked with the test group chair if there is subjective improvement for at least one more sequence (except Race) - the result was positive.

Adopt to JMVM.

182

53.1.1.1.2 JVT-W023 ( Info) [S.-C. Lim, D.-H. Han, Y.-L. Lee] CE5: Verification of loop filtering in MVC

This document presents verification results of “CE5: Loop filter” proposal by Samsung and KHU. The encoder and decoder executables, bitstreams, source code, and configure files were provided by Samsung and KHU. And the provided source code was compiled and the decoder executable was run with the provided bitstreams. All of the decoded results were reported to be matched exactly with the results provided by Samsung and KHU.

Checked with the same source code, results verified.

53.1.1.1.3 JVT-W031 ( Prop 2.2) [J.-H. Yang] CE5: Illumination comp. info. derivation for MVC

This contribution proposes the modification of the part “2.3.2 ICA MC for Skip and Direct modes” in JMVM 3.0. Unlike the P_Skip mode, the B_Skip mode in the current JMVM model requires the transmission of mb_ic_flag and dpcm_of_dvic. The proposed scheme derives the IC information (mb_ic_flag and DVIC) from the neighboring blocks for the B_Skip mode. Then, it is asserted that the IC technology with the proposed B_Skip mode requires simple syntax and becomes in line with the H.264/AVC design. Also, the simulation results are asserted to show that the proposed scheme gives better coding efficiency

Revisit of JVT-V063, implementation on the newest JMVM version was done. B_direct mode case of JVT-V063 not used any more, B_skip mode case is retained. Derivation of IC done similar to the B_skip mode for motion vectors of AVC. Proposal for syntax modification relative to JMVM 3: Remove syntax elements from slice header and macroblock prediction syntax. 0.4% bitrate saving on average.

Performed further study on possible complexity impact in breakout, reported back. Complexity decreased without penalty in compression.

JVT decision: Adopt to JMVM.

53.1.1.1.4 JVT-W085 ( Info) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] CE5: Verification of JVT-W031 illumination comp. info. derivation

In this document, verification results of JVT-W031 are reported. Reports that they received the source code, configuration files, coded bitstream, experimental results, and documents for description. LG proposed the derivation scheme for IC information in B_Skip mode. They verified the implementation, encoding/decoding for the proposed scheme and its results.

Checked with the same source code, results reportedly verified.

53.2 CE 6 & related docs: MVC view interpolation

53.2.1.1.1 JVT-W055 ( Prop 2.2/3.1) [T. Senoh, M. Okui, K. Enami] Experimental results of camera-rotation-compensated prediction in CE6

Experimental results of view interpolation prediction based on the camera-rotation-compensation of the reference pictures were reported. For Uli test sequences, very small R-D gain was reportedly observed. The reasons reportedly seem to include camera location errors, camera gain errors and many occluded blocks.

183

(no verification)Only small gain found (<0.1 dB). No further study envisaged currently, but might be combined with block slant distortion compensation. Contribution noted.

53.2.1.1.2 JVT-W059 ( Prop 2.2/3.1) [S. Yea, A. Vetro] CE6: View synthesis prediction

This contribution reports progress of CE6 on view synthesis prediction for multiview coding. A method to determine an appropriate depth search range and step size has been explored. It is asserted that these parameters have a substantial effect on the coding gains. Improved coding results are shown for one test sequence, however overall gains for other sequences are negligible. The report suggests that further study is needed to determine the appropriate depth maps for each test sequence.

(no verification)Finding depth range and step size by using KLT tracker. Correction vectors used in addition to depth. Adaptive strategies aiming to reduce the coding cost, also depth range may change temporally over the sequence. Breakdancer bitrate saving up to 8% (only at low bit rates as high), not much gain for other sequences currently. Plan further study in particular for improved depth search and adaptive coding, and alternative representations of depth. Contribution noted.

53.2.1.1.3 JVT-W084 ( Info 2.2/3.1) [Y. S. Ho, K. J. Oh, C. Lee, B. H. Choi, J. H. Park] Observations of multi-view test sequences

This document introduces information obtained by observing the multi-view test sequences. The observations are related to vertical and horizontal displacement caused by inaccurate camera arrangements, illumination changes, synchronization of multi-view sequences, and focusing. Future multi-view video sequences should solve these problems for efficient multi-view coding and real applications.

Reported as introduction before JVT-W083. Reports problems: Vertical displacements, illumination changes, synchronism, camera arrangement (in case of Rena sequence). Contribution noted (may be difficult to get better test sequences).

53.2.1.1.4 JVT-W083 ( Prop 2.2/3.1) [Y. S. Ho, C. Lee, K. J. Oh, B. H. Choi, J. H. Park] CE6: View interp pred for MVC

This contribution describes a ‘VIP P-picture’ coding which uses the synthesized image as the additional reference frame. The proposed view interpolation method can make an intermediate image by using initial disparity estimation, variable block-based disparity estimation, and pixel-level disparity estimation based on the adjusted search range. In addition, motion vector prediction scheme is modified and vertical displacement is compensated to maximize the efficiency of ‘VIP P-picture’ coding.

Try to compensate the problems reported in JVT-W084 (in particular vertical displacement compensated before disparity estimation). Modified motion vector prediction in cases where neighboring blocks are mixtures of VIP and V/T frames. For “dense sequences” (Akko&Kayo and Rena) average gains 0.2 dB overall, 0.66 dB for B-views, for other sequences marginal (Breakdancers) or no gains. (In general, the rate for B-views is not too high anyway.)

No action taken.

184

53.2.1.1.5 JVT-W103 ( Info) [J.-H. Yang, S.-H. Lee] CE6: Verif GIST MVC contribution JVT-W083 MVC view interp pred

This document reports verification results of JVT-W083 from GIST. The author received decoding executables, coded bitstreams for B-views, reconstructed yuv files and experimental results. They verified the decoding and its results for the proposed scheme.

Checked with the same source code (source or executable? document says executable), results verified for 3 sequences where gain was observed.

53.2.1.1.6 JVT-W096 ( Prop 2.2/3.1) [S. Naito, A. Koike] CE6: Results on MVCIn this contribution, recent progress for CE6 on view interpolation prediction for multiview video coding is described. In order to improve the coding efficiency by view interpolation and disparity compensation, an efficient encoding scheme for depth and disparity vectors is proposed. Experimental results for anchor frames are provided under the common test conditions. The proposed scheme is asserted to be effective especially for a sequence with an arc camera arrangement.

(no verification)Introduce a coding scheme for disparity vector and depth. Basic idea to allow conversion between depth and disparity vector, in order to use them mutually for prediction from neighboring blocks. Depth is derived using camera parameters plus disparity information (on block basis). Maximum gain reported for Breakdancer (0.1..0.2), almost nothing for other sequences. In Breakdancers, difference between depth and disparity vector is apparently most significant.

No action taken.

53.2.1.1.7 JVT-W087 ( Prop 2.2/3.1) [S. Shimizu, H. Kimata] New view synthesis pred framework using resid pred

This contribution proposes a view synthesis prediction framework for multiview video coding using residual prediction. In this framework, only one depth map is encoded at every instant in order to perform view synthesis prediction with fewer bits as a whole. It is asserted that the most important technique proposed in this contribution is spatial/temporal residual prediction on view synthesis prediction residual signals. The preliminary experimental result for the sequence “rena” was reported as -7.39% or 0.34 dB in Bjontegaard measure. Note that this experiment was conducted on the special prediction structure for low delay.

(new contribution)Focus on issues: How to reduce bits for depth information, how to deal with inaccuracies in depth estimation. Depth is encoded duplicated (e.g. as disparity for different pictures), in fact being redundant due to same physical meaning. Goal to encode only one overall depth map (e.g. on base view) and derive all other reference information out of it. Residual signal between original and depth-synthesis prediction is encoded, but may have spatial and temporal correlation (because the wrong camera parameters do not change over time). Therefore, they can be encoded like conventional video after depth-based synthesis prediction from the base view. Average gain for Rena (no other sequences tested) 0.34 dB average. No temporal prediction applied for the non-base views.Looks interesting (also some relation with depth-based projection in 23002-3).

185

Current conclusion from CE: Breakout to identify the most promising directions in view interpolation and start more collaborative effort.Currently, most gains are reported for Rena Akko&Kayo and Breakdancers. For other sequences, due to the physical structure the ranges of depth are much to high to be compensated by the global camera parameters.

Schemes that estimate depth at decoder are not followed any more currently. Would only work for dense sequences.

53.2.1.1.8 Anthony Vetro presents new CE6 work plan.Two different paths so far: Block-based depth, pixel based depth (mostly from global camera parameters) – latter does not work for sequences with highly varying depth– do not further follow decoder-side depth estmation– concentrate on approaches video plus depth– Two approaches: Directly coded residual, predictively coded residualIssues:– What is resolution, range and precision of depth maps– Study global depth, try to minimize rate for depth mapsCurrently aiming for improved coding efficiency, but would be interesting to study the relationship with the video plus depth approaches that were presented (for view synthesis). Tradeoff: The latter one would require more precise depth maps which might penalize the compression performance. Needs to be further studied. Continue CE. Discussed combination of different approaches available so far in joint software framework. Uploaded slides as JVT-W133.

53.2.1.1.9 JVT-W133 -B (BoG) [A. Vetro] BoG report on MVC view interpolation pred

Summary of BoG Discussion on View Interpolation Prediction.

53.3 MVC high-level syntax

53.3.1.1.1 JVT-W035 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] On MVC JD 2.0

This contribution presents some comments and proposals on the following topics: 1) single-path adaptations based on priority and temporal level, 2) view_level and indication of suffix NAL unit, 3) IDR picture, IDR access unit, changing and activation of sequence parameter sets, 4) implicit removal of decoded non-reference pictures that belong to the not-output views, and 5) scalable nesting SEI message.

Issue 1 (Priority ID): JVT decision: Proposal adoptedIssue 2 (view_level): align with the SVC design (decided not to have suffix NAL unit, see JVT-W125). View level cannot directly be compared with temporal level, because there are much more different configurations. In principle, no semantics is associated. Offline clarification resulted in a recommendation to remove view_level. JVT decision: Agreed.Issue 3 (IDR, IDR access, SPS)

a) IDR, can other pictures in same AU be non V-IDR? JVT decision: Proposal adopted.b) Shall IDR access unit have all pictures IDR or V-IDR (but we may need a name for this

case - editorial)186

c) When can SPS change? Only in IDR – JVT decision: Agreed.d) Shall SPS MVC extension be same for all SPSs? Comment: View dependency should be

retained the same, otherwise start with new IDR. JVT decision: Adoptede) What happens if certain views are stripped off, but SPS is unchanged? Should there be an

identifier for discardable views in SPS? Other solution could be to signal this by SEI message (as in SVC) – thin about in future. No action.

Issue 4 (implicit removal of decoded pictures) editorial – clarify offline.Issue 5 (re-use scalability nesting SEI message in backward-compatible manner) JVT decision: Agreed, but may need re-consideration / extension when the views shall be differently scaled temporally, spatially.

53.3.1.1.2 JVT-W036 ( Prop 2.2.1/3.1) [Y.-K. Wang, M. M. Hannuksela, Y. Chen] MVC output related conformance

MVC supports a large range of views, but the number of the views the decoder process can be constrained to a relatively small value to meet the rendering capabilities, for example. According to the current MVC draft, it cannot be known from the bitstream which views are to be outputted. It is claimed in this contribution that the information which views are to be outputted is required in the picture output and removal processes of the hypothetical reference decoder as well as in the derivation of the minimum decoded picture buffer requirement. While it is possible for a decoder to get the information through a systems means that is out of the scope of the MVC specification, it is asserted that containing the information within the bitstream is helpful in at least two aspects. First, like AVC or SVC, the decoding process can be independent of external information. Second, when parts of the bitstream have not been received due to any reason, the receiver knows how to handle, e.g. to conceal a lost picture or to omit decoding a non-required picture. This contribution proposes the signaling of the to-be-outputted views within MVC bitstreams.

One possibility to leave this unspecified. However, if there is a mechanism to specify this, it may even be possible not even to decode these views. Sounds like a very special case, where e.g. the server or proxy must be aware of the type of display that is available at the receiver end.

Discussed further after offline of more showcase details with Anthony V.

JVT decision: Adopted into SEI. Showcase to be made by next meeting.

Remark: Using SEI for this seems odd, since it governs normative decoder behavior.

53.3.1.1.3 JVT-W037 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] View scalable SEI message for MVC

View scalability information SEI message is proposed in this contribution. The SEI message is used to signal a number of operation points and information of each operation point, including definition of the operation point, maximum priority_id and temporal level, profile and level compatibility information, bitrate information, frame rate information, and initial parameter sets information.

Similar to scalability information SEI message in SVC. JVT decision: Adopt.

53.3.1.1.4 JVT-W038 ( Prop 2.2.1/3.1) [Y. Chen, Y.-K. Wang, M. M. Hannuksela] Operation point and view dependency changes SEI messages for MVC

View scalability information SEI message is proposed in this contribution. The SEI message is used to signal a number of operation points and information of each operation point, including

187

definition of the operation point, maximum priority_id and temporal level, profile and level compatibility information, bitrate information, frame rate information, and initial parameter sets information.

View dependency and scalability operation point changes cover very specific case – keep this for further study, no adoption in the current early phase of the project.

53.3.1.1.5 JVT-W039 ( Prop 2.2.1/3.1) [Y.-K. Wang, Y. Chen, M. M. Hannuksela] Non-required pictures SEI message for MVC

A new SEI message for indication of non-required pictures is proposed in this contribution. With the proposed SEI message, a communication system using MVC can avoid transmitting, decoding and buffering of the non-required pictures. A non-required picture refers to such a picture in a certain view in an access unit that is not used for inter-view prediction while listed as an inter-view prediction picture in the sequence parameter set. Furthermore, a non-required picture does not affect the decoding process of the current and future pictures in the current view and other target output views.

See notes in section on JVT-W056.

53.3.1.1.6 JVT-W056 ( Prop 2.2) [J. B. Choi, W. S. Shim, H. S. Song, Y. H. Moon] Inter-view prediction reference picture marking

This document proposes to additional nal_ref_idc_view syntax for the marking process of inter-view prediction reference picture and the modified initialization process for reference picture list for inter-view prediction picture. When the prediction structure that some pictures are used for inter-view prediction reference and some pictures are not used for inter-view prediction reference in same view is implemented, it is asserted that current inter-view prediction reference picture marking system has some problems. Firstly, current inter-view prediction reference picture marking system could mark a picture which is used for inter-view prediction reference picture. Actually the picture is not used for inter-view prediction reference picture. Because current inter-view prediction reference picture marking system only uses view dependency information from SPS. Secondly, current initialization process for inter-view prediction reference picture could insert the picture that is not used for inter-view prediction reference in reference list. Because Current initialization process considers the view dependency information and PicOrderCnt(). The proposed nal_ref_idc_view represents whether a picture is used for inter-view reference picture and the modified initialization process considers the view dependency information, PicOrderCnt() and Proposed nal_ref_idc_view.

Case 1: Some pictures in same view are not use for interview prediction, case 2: In case of

temporal resolution of some views is different. Similar method was proposed in U103. Necessary information can be derived from view dependency id.

Consensus in the group that JVT-W039 and JVT-W056 cover something useful. Breakout group to elaborate on unified solution (considering pro’s and con’s of doing it in NAL or as SEI, also relevance in terms of complexity saving) and report back.

After consideration – JVT decision: Adopt JVT-W056 (not JVT-W039 at this point).

53.3.1.1.7 JVT-W066 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Ref pic list reordering for MVC

188

In the current JD, new Reference Picture List Reordering (RPLR) commands were added to support reordering of inter-view reference pictures. This document proposes to change the equations used to derive the view index prediction value in order to allow for duplicating/repeating the inter-view reference pictures in the list.

Resolves problem that currently exists for the first RPLR command. JVT decision: Adopt. Remark: There are surely other mechanisms to fix the problem, but the proposed method is similarly simple as those would be.

53.3.1.1.8 JVT-W067 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] H.264/AVC extension for MVC using SEI message

This document proposes a new supplemental enhancement information (SEI) message for signaling of multiview information in a H.264/MPEG-4 AVC compatible bitstream where each picture contains sub-pictures for each particular view. This SEI message is intended for easy and convenient display of multiview video streams on 3D monitors which may use such a framework.

Reports a method that would enable packing several views into an AVC compatible bitstream. (using tiling of views).

There would be other possible approaches to achieve this (e.g. temporal interleaving). Could also be seen as extension of stereoscopic SEI. Anyway, existing AVC decoders would be unaware of such a new SEI message, which could only be defined in a new amendment. Set up AHG on “Study of MVC solutions using existing AVC decoders”, chair P. Pandit.

53.3.1.1.9 JVT-W074 ( Prop 2.2) [H. S. Song, W. S. Shim, Y. H. Moon, J. B. Choi] Comments on view dependency info

This contribution consists of two sections. The section 1 is about the additional syntax of sequence parameter set for the flexibility of the inter-view prediction structure. Since the added new syntax gives information which a picture with the temporal level is predicted with the inter-view prediction, the proposed scheme is asserted to be efficient for the random access or memory management, etc. Also, the proposed scheme can reportedly be used under the environment with the restricted memory size or the required low complexity.

The section 2 is about the modified representation method of view dependency information for efficient representation view dependency. The modified representation method represents view dependency by the basic unit of repeat pattern and the number of view of basic unit. It is reported to be useful and efficient for representing prediction structure which the view dependency is repeated by uniform pattern.

Comment on first part: Seems not to be very significant in terms of saving memory and complexity. No real support in the group.

Comment on second part: Amount of bits saved is negligible and not worth the additional complexity.

No action taken.

53.3.1.1.10 JVT-W080 ( Prop 2.2.1) [K. Ugur, H. Liu, Y.-K. Wang] Showcase for parallel decoding info SEI message for MVC

189

At the Marrakech meeting, the Parallel Decoding Information SEI message (JVT-V098) was adopted to JMVM to facilitate parallel encoding/decoding of different views. This contribution presents a showcase for this SEI message. In addition, some allegedly-minor issues were identified with the syntax and semantics of the SEI message after the Marrakech meeting. This contribution also proposes the changes to syntax and semantics to address these issues.

Group is satisfied with showcase. JVT decision: Agree with syntax adjustments as presented.

53.3.1.1.11 JVT-W088 ( Prop 2.2) [S. Lin, S. Gao, Y. Liu, L. Xiong] H.264/AVC SEI extensions for MVC

This contribution proposes modifications to extend two H264/AVC SEI messages for MVC, one is spare picture SEI message, the other is decoded reference picture marking repetition SEI message. Both of these SEI messages were introduced in H.264/AVC to implement error concealment.

Remark: Can achieve same functionality with JVT-W035.

Response from proponent: That does not enable inter-view spare picture selection. Suggests that spare picture usage in the view direction would be useful.

Question: Any example pictures identified where this would be useful?

Remark: Marking process only affects the temporal direction. There is no marking that operates in the view direction.

Aspects of this contribution beyond what can be achieved by JVT-W035 are for further study.

53.4 MVC other normative technical inputs

53.4.1 MVC motion/disparity vector coding

53.4.1.1.1 JVT-W081 ( Prop 2.2) [H. S. Koo, Y. J. Jeon, B. M. Jeon] MVC motion skip mode

This document proposes a motion skip mode for MVC which is originated from the idea that there is a similarity in respect of motion between the neighboring two views. In the proposed method, the motion information is inferred from the corresponding macroblock in the frame with the same temporal index of the neighboring view. To compensate the inter-view difference generated by camera geometry, disparity vector is applied to find the corresponding macroblock in the neighboring view. The maximum gain obtained with the proposed method is up to 0.54 dB.

Uses global disparity vector, for non-anchor pictures this is derived from the anchor pictures. Introduces the global disparity in the slice header syntax. Introduce motion_skip_flag in MB layer syntax. 0.54 dB gain for Rena, 0.38 Akko&Kayo, Race 0.25, Flamenco 0.1, negligible for other sequences. 0.18 dB on average.

Comment: Proposal uses inter-view reference for motion information in view level 1, which would require to define the picture as reference picture and store it in DPB. Not clear if this is possible.

190

Clarified buffer management issue offline and reported back. After review of breakout: Proponent was to produce concrete description text for potential inclusion in JMVM – participants of breakout also were asked to check this against the JMVM software code.

Break-out group discussion held with results recorded in JVT-W139.

53.4.1.1.2 JVT-W139 -B (BoG) [LG, Thomson] Break-out conclusions on JVT-W081Report of break-out discussion on JVT-W081.

JVT decision: Adopt (into JMVM) as recorded in JVT-W139.

53.4.1.1.3 JVT-W073 ( Info) [K. Sohn, J. Seo] Verification of JVT-W081 LGE MVC motion skip contrib.

This document reports the cross-check results of JVT-W081 “MVC motion skip mode” by LGE. The source code, configuration files and coded bitstreams were provided. The verification was performed by decoding the bitstreams provided by LGE. The simulation results of JVT-W081 are confirmed.

Check made on basis of compiled source code, results verified.

53.4.1.1.4 JVT-W101 ( Prop 2.2) [H. Yan, J. Huo, Y. Chang, S. Lin, S. Gao, L. Xiong] MV/DV prediction based on RDV

This document is a response to JVT documents JVT-V071, JVT-V072 and JVT-V073. Several changes of original techniques have been made, and coding performance of proposed mv/dv prediction method is investigated.

Only small or no gains. Possibly still bugs in implementation. Contribution noted.

53.4.1.1.5 JVT-W104 ( Prop 2.2) [S.-H. Lee, S.-H. Lee, N.-I. Cho, J.-H. Yang] MVC disparity vector pred

This contribution is a response to the ad hoc group work on disparity and motion vector coding. We propose a modified motion vector prediction scheme, which distinguishes neighboring motion vectors as temporal motion vectors and disparity vectors. Each kind of motion vector is used exclusively in motion vector prediction phase by reference picture types. Disparity vectors are derived from temporal matching blocks when they are not available from neighboring blocks. Proposed algorithm shows 0.0 dB~0.04 dB PSNR gain and 0.2%~1.2% bit reduction with Bjontegaard measure for all views and all frames. And 0.01 dB~0.074 dB PSNR gain and 0.34 % ~ 2.76% bit reduction for selected views which have an inter-view dependency in non-anchor frame.

Average gain without RPLR on: 0.042 dB, with RPLR on: 0.006 dB. In some cases worse results than JMVM 3.0.2. Contribution noted.

53.4.1.1.6 JVT-W107 ( Info) [K. Sohn, J. Seo] Verif JVT-W104 MVC disparity vector pred

191

This document reports the cross-check results of JVT-W104 “MVC disparity vector prediction” by SNU/LGE. The binary files, coded bitstreams were provided. The verification was performed by decoding the bitstreams provided by SNU/LGE. The simulation results of JVT-W104 are confirmed.

Cross-check based on compiled source code.

53.4.2 MVC weighted prediction

53.4.2.1.1 JVT-W040 ( Prop 2.2.1/3.1) [S. Liu, Y. Chen, Y.-K. Wang, M. M. Hannuksela] Constraints on temporal direct mode and weighted prediction in MVC

When an inter-view prediction picture is a co-located picture, it is reportedly not specified how the temporal direct mode and implicit weighted prediction should be applied. It was studied whether the temporal direct mode suits the inter-view prediction pictures by using view_id instead of PicOrderCnt to calculate the scaling factors. Judging from the simulation results, the modified temporal direct is reported to provide no efficiency gain and sometimes to even bring coding efficiency loss. It is therefore proposed that when the co-located reference picture belongs to inter-view reference pictures, temporal direct mode shall not be used. Furthermore, methods on how to support weighted prediction when there are inter-view references in the reference lists are discussed.

Disable temporal direct mode in case of inter-view prediction. JVT decision: Adopt. Problem with scaling in case of implicit weighted prediction. JVT decision: Adopt solution to disable implicit weighting prediction (which seems to be the best possible fix for the time being).

53.4.2.1.2 JVT-W098 ( Prop 2.2) [J. H. Park, Y. H. Kim, J. W. Kim, B. H. Choi] Weighted prediction for MVC

This contribution suggests re-use of base view weighting parameters when all view sequences have quite similar tendency of weighting parameters. It reports that introducing one bit syntax same_weighted_prediction_flag in SPS in SVC MVC extension gives a way to avoid redundant process when all views have the same weighting value. Also reported is that introducing use_base_view_prediction_flag in slice header gives flexibility.

Proposal to re-use the weighting parameters from base view for enhancement view. Coding gain negligible, but would need change of existing slice header syntax and decoding process. Contribution noted.

53.4.3 MVC downsampled reference etc.

53.4.3.1.1 JVT-W079 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Inter-view prediction with downsampled reference pictures

In this document is proposed a new inter-view prediction method in case that each view has a different spatial resolution in a multi-view sequence. The above case is asserted to be beneficial to reduce both the total bit rate for multi-view sequence and the complexity of encoding and decoding multi-view sequences, because the number of samples in some views could be decreased when spatial resolution of them is decreased. The contribution discusses the efficiency

192

in terms of coding efficiency and complexity of the proposed coding in the document. And in the proposed method, it is asserted that we could have an optional post processing in which decoded images of such low resolution pictures could be up-sampled. It also discusses briefly the effectiveness of such post processing.

Idea to reduce the spatial resolution in some of the views to achieve better compression performance. This could also include the case where some of the cameras produce less resolution. Preferably, these are encoded as B views. Requires downsampling in the prediction process for the B views. Proposal to add the respective syntax in the SPS. Currently, it would only be possible to encode the downsampled views independently from the higher-resolution views. Results indicate that significant compression gain can be achieved for the low-resolution views if prediction from the higher-resolution views is enabled.

Comments:– In practice (for certain types of displays) also inter-view methods would be used to generate

the up-sampling– Complexity reduction is very interesting aspect– Overall rate saving needs still to be investigated (currently, only saving on B views was

reported)– Would also require subjective evaluation

Further work encouraged (see below under JVT-W092)

53.4.3.1.2 JVT-W092 ( Prop 2.2/3.1) [P. Pandit, P. Yin, C. Gomila] Reduced resolution update for MVC

This document presents the extension of Reduced Resolution Update (RRU) mode for multi-view video coding (MVC). RRU is currently supported by H.263 (Annex U). This mode is asserted to provide the opportunity to increase the coding picture rate while maintaining sufficient subjective quality. This is done by encoding an image at a reduced resolution, while performing prediction using a high resolution reference. This reportedly allows the final image to be reconstructed at full resolution and with good quality, although the bit rate required for encoding the image has been reduced considerably. It is asserted that the results using JM 10.1 show performance improvements of about 0.3 db over not using RRU.

Request for more flexibility in having views of different resolutions, similar to JVT-W076. Support for spatial scalability. Preliminary results with RRU (based on JM, not JSVM), gain of 0.2 … 0.6 dB for the views that were processed (no overall gain reported). Compared to the previous proposal, it is here necessary to specify a normative upsampling filter in the loop. Saving in complexity not as high as 076 (only for parsing process it is less, but all MC/DC is done at full resolution).

Further study of this and JVT-W076, establish AHG, still need to be checked whether there is relationship with other issues such as view interpolation.

53.4.3.1.3 JVT-W094 ( Info 2.2) [W. J. Tam] Image and depth quality of asymmetrically coded stereoscopic video for 3D-TV

This contribution provides experimental results of subjective evaluation of stereoscopic images in which the image quality of the left-eye and the right-eye views are different. Subjective ratings of binocular image quality are biased towards the input with the higher image quality when the source of image degradation is Gaussian filtering (blur). When the source of image degradation

193

is from quantization, such as from block-based coding, the binocular image quality is reported to be approximately the average of the image quality inputted to the two eyes. In contrast to image quality, depth quality ratings were reportedly only slightly affected by asymmetrical image degradation arising from either blur or blockiness. The main findings were replicated for a wide range of asymmetrical quality between the two eyes, using a variety of test sequences, and for different groups of viewers.

Finding that lowpass-filtered image for one eye gives subjectively same quality perception in stereo if resolution for the second eye is kept high. Asymmetrical coding is viable method for bandwidth saving. Cross-switching at scene cuts is uncritical and might help to circumwent problems of large asymmetries.

Would need more investigation how it works for multi-view; potentially alternating quality (low-high-low-high) could be viable. In general, this is more an issue of encoder decisions. Contribution gives valuable hints, but difficult to quantify in absolute numbers (most probably sequence dependent)

53.4.4 MVC modes and other coding efficiency topics

53.4.4.1.1 JVT-W078 ( Prop 2.2/3.1) [H. Kimata, S. Shimizu, K. Kamikura, Y. Yashima] Co-located block condition for inter-view prediction

This is a follow-up of JVT-U134, in which a simplified coding method of direct mode for inter-view prediction of MVC was proposed to reduce memory usage for disparity information. In this document, all coding results are presented and complexity reduction is discussed.

Main issue is reducing memory bandwidth for disparity vectors. Usage of co-located block meaningless for inter-view prediction, shown that there is no loss in compression performance. Not clear how large the complexity reduction really is, most probably it is minor. Therefore, it seems to be better to keep it as it is (as in AVC) instead of imposing additional constraints that are specific for MVC.

53.4.4.1.2 JVT-W082 ( Prop 2.2) [Y. J. Jeon, H. S. Koo, B. M. Jeon] Modified spatial direct mode in MVC

The application of the colZeroFlag derivation process in spatial direct mode is not proper in some cases where the motion properties of current block and co-located block are different (i.e. one is mv, and the other is dv, or vice versa). This contribution introduces a solution to keep this derivation process meaningful. Before the derivation of this flag, a validation check process is invoked to check whether the motion properties of two blocks are identical. If the motion properties of two blocks are identical, the existing colZeroFlag derivation process is invoked. If not identical, colZeroFlag is set to 0 without any further investigation. By the proposed method, the colZeroFlag derivation process in H.264/AVC can be carried out in proper way.

Zero gain in compression, no reduction in complexity, keep it as it is in AVC. Contribution noted.

53.4.4.1.3 JVT-W065 ( Prop 2.2/3.1) [P. L. Lai, A. Ortega, P. Pandit, P. Yin, C. Gomila] Adaptive reference filtering for MVC

194

This document considers the problem of coding multi-view video that exhibits mismatches in frames from different views. Such mismatches could be caused by heterogeneous cameras and/or different shooting positions of the cameras. In particular, it considers focus mismatches across views, i.e., such that different portions of a video frame can undergo different blurriness/sharpness changes with respect to the corresponding areas in frames from the other views. It proposes an adaptive filtering approach for inter-view prediction in multi-view video coding. Preliminary results, on anchor only coding (IPPPP), are asserted to show gains ranging from 0.06db to 0.8db over the current method. The asserted gain is larger for sequence with stronger focus mismatches.

Gains depending on number of reference pictures that are used. For one sequence (Flamenco 2) gain of > 1 dB and almost 20% bitrate savings are reported (for selected views). Decoder complexity increased by a 5x5 2D filter. Average gain 0.45 dB for case of 1 reference picture, 0.14 for 3 reference pictures, 0.06 for 5 reference pictures. Further study in AHG, in particular consider complexity at pixel level, possibly combination with subpel interp. Filters (has some relation with VCEG AHG that exists for studying adaptive MC interpolation filters and also with previous proposals (Wedi) for adaptive Wiener loop filters).

53.4.5 MVC depth-based methods & displays

53.4.5.1.1 JVT-W095 ( Info 2.2) [W. J. Tam, L. Zhang] Depth map preproc and minimal content for 3D-TV using depth-based rendering

This contribution provides experimental results of subjective evaluation of stereoscopic images consisting of an original image and a rendered image from depth-image-based rendering (DIBR). Experimental results show the beneficial effect of smoothing of depth maps before DIBR on image quality. Furthermore, results are shown for asymmetrical smoothing in which the extent of smoothing is larger in the vertical than in the horizontal direction to reduce geometric distortions. Finally, consistent with the findings that depth maps do not have to contain "full resolution," subjective assessment results from a different set of studies indicate that enhanced depth sensation, compared to reference monoscopic images, can be obtained using "surrogate" depth maps. That is, depth maps that contain sparse "depth" information located mainly at edges and object boundaries. The overall findings indicate that depth information for DIBR, just as for colour information, do not have to be of full spatial resolution for the generation of useful images for autostereoscopic multiview displays and other stereoscopic displays to produce enhanced sensation of depth.

In general indication that stronger smoothing of depth maps provides improved subjective quality. Depth of boundary location may be sufficient. Right view generated by depth-based projection from the left. Subjective tests performed with minimum of 10 subjects. Shutter-eye glasses used.

53.4.5.1.2 JVT-W100 ( Prop 2.0/3.1) [A. Smolic, K. Mueller, P. Merkle, N. Atzpadin, C. Fehn, M. Mueller, O. Schreer, R. Tanger, P. Kauff, T. Wiegand, T. Balogh, Z. Megyesi, A. Barsi] Multi-view video plus depth (MVD) format for advanced 3D video systems

The contribution proposes to initiate a study on how to support multi-view video plus depth (MVD) data efficiently by a coding standard. It illustrates advanced 3D video and free viewpoint video systems, and argues that these are not efficiently supported by available and emerging specifications, such as MPEG-C Part 3 and MVC. The central requirement of such technology is

195

said to be an input data format that allows rendering a wide range continuum of views at the decoder. MVD is introduced and illustrated in some detail, being multi-view video with multiple associated per sample depth maps. It is claimed that MVD fulfills the above requirement and is therefore a suitable candidate for a basic format for advanced future 3DV and FVV systems. Finally, an initial work plan for the proposed investigation is presented.

Relationship with both MVC and MPEG-C part 3. Required input format that allows rendering of continuous views. Occlusions can be handled by smoothed depth maps. Single video plus depth has limitations, artifacts when wide range is required. Possible solution is multiple videos plus depth, another is layered depth video: One video, one depth map, one background layer for occluded pixels. Proposal to start new work plan on this. Firstly, this is about new functionality. The relationship with compression would also be interesting to be investigated (e.g. using depth maps for generation of prediction references). Report that it was found that compression of depth maps for multiple views is not simple (if high quality view generation is required).

Look into issues of– Compression of depth maps– Relationship with MPEG-C part 3– Relationship depth-based rendering and view compression (CE 6)

Was further discussed in the context of requirements (FTV).

Discussion (Tue morning): Is it necessary to define normative rendering? Most raise objections against that. Displays are that specific that it needs to be left to the manufacturer how to perform the interpolation. It must be specified what the “conformant” output views are (may not be equal to the views that are actually captured). Definition of data representation that allows to generate a certain number (in principle up to arbitrary) of views. Boundary between decoding and rendering may be floating, depending on whether a method for rendering would be worthwhile to be considered as a compression tool (e.g. producing a better prediction of intermediate views).Under discussion:– Format allowing generation of arbitrary (up to continuous) views would be useful,

supporting many types of displays (consensus on this)– Would require (in addition what is currently investigated) to have information about 3D

scene structure as necessary for rendering (one example would be depth maps)– Rendering/display/interpolation (see note below) is non-normative, but an example method

shall be given, and would be needed anyway for the development– Needs to be investigated whether relationship between depth information and picture

information helps to develop a better compression

Note: There is some internal dispute on what “rendering” means. Interpolation may also include spatial upsampling in cases where some of the views have lower resolution.

Further discussed with Requirements (Wed 14:00), also whether this will be added into current MVC development or another activity with extended timeline.

How to evaluate effectiveness of depth maps as a view coding feature?

Question raised by JVT-W100: Should we specify normative interpolative rendering? Without it, how does an encoder know how to optimize its encoding decisions?

Remark: Leave that non-normative.

196

Remark: Would like at least some decoders to be required only to extract and decode exact (non-interpolated) view(s).

See also related notes in section on JVT-W127.

Further study to be held in CE on view interpolation prediction.

53.4.5.1.3 JVT-W060 ( Prop 2.2/3.1) [A. Vetro, S. Yea, W. Matusik, H. Pfister, M. Zwicker] Anti-aliasing for 3D displays

This contribution describes an anti-aliasing technique for improved rendering of multiview video on 3D displays. View interpolation techniques are utilized to achieve an oversampling of the multiview signal in the view dimension. The oversampled signal is then filtered to suppress high frequency portions of the signal that contribute to aliasing, and finally sub-sampled to match the display characteristics. This contribution examines ways to minimizes receiver resources in this framework. Two distinct needs for MVC are highlighted, including the need to code and transmit depth maps along with the multiview video, as well as the need for spatial scalability. An SEI message that signals acquisition and scene attributes is also proposed.

Danger of alias: Relationship between spatial resolution and number of views (including scalablity of both). Spatial resolution of each view affects the spectrum of input signal. One effect is ghosting artifacts which can be prevented by pre-filtering. Depth maps could be part of access unit and managed together with the primary picture set (which would not be possible when using MPEG-C part 3).

Proposal for maximum disparity and camera parameters as SEI messages. The group further discussed the proposed SEI message issues in the context of HL syntax. 1) For camera parameters, participants were asked to clarify relationship with previous proposals, precision etc., 2) for max. disparity showcase is needed.

JVT decision: Adopt camera parameters SEI and max disparity SEI (showcase expected at next meeting).

Min disparity may also be useful – for further study.

53.4.6 MVC view parallel processing

53.4.6.1.1 JVT-W077 ( Prop 2.2) [P. Yang, X. Xu, G. Zhu, Y. He] View parallel processing on MVC

A view parallel coding architecture is presented in this proposal. In MVC inter-view references are used to exploit the dependency of different views. Consequently, the parallel processing ability is deteriorated. When parallel processing is required, a simple structure would be used. However, the coding efficiency for this scheme is therefore compromised. The proposed method in this document restrains the prediction between pictures at the same time slot in different views, but allows the other kinds of inter-view prediction, thus all of the views can be processed in parallel. Therefore, the proposed method has a similar parallel mechanism as the simple structure, while achieving some coding efficiency gain over it. An average PSNR gain (for all encoded pictures) is reported to be 0.14 dB when encoding the common test conditions and 0.22 dB for all non-key frames. Also, the proposed method would favorite sequences with large motions. An overall 0.34 dB gain and 0.47 dB gain for non-key frames are achieved.

197

Implementation done in JSVM (JMVM only supports view-first). 0.14 dB gain on average as compared to “simple” structure (which does not allow inter-view prediction for non-key pictures). Concept may have implications on access unit definitions and buffer management. Encoding of frame t0 would cause initial delay which can never be catched up again. Identify relationship with JVC proposal for parallel processing made 2 meetings ago.

JMVM implementation would be needed.

No report given Tuesday morning, apparently no offline discussion happened prior to that time. Cross-check JVT-W108 was still not available at that time.

Remark: Complicates management of reference pictures. Would like to see text on how this would be solved. We have cross-view and cross-time dependency referencing – this proposes new diagonal dependency directions.

Remark: Implications on MMCO and ref pic list construction may be major.

Remark: Consider JVT-V132 structure. IPPP cross-view from that document is suggested as a better reference. Commenter asserts that non-hierarchical structure in view direction will provide better results than hierarchical structure.

For further study.

53.4.6.1.2 JVT-W108 -QV (Late Info) [Q. Chen, Z. Chen] Verif JVT-W077 view parallel proc on MVC

This document verifies the results of JVT-W077: “View parallel processing on MVC” from Tsinghua Univ.

Based on the executable files (encoder and decoder) and configure files provided, bit streams were reportedly generated for verification.

In the first (late) uploaded version, only the sequences race and exit were reported finished. The finished results were reportedly identical to JVT-W077r1.xls and can be found in JVT-W108.xls. The bit streams could reportedly be decoded correctly.

Remark: This “verification” does not seem to fulfill the spirit of such efforts – the algorithm was not investigated, and the contribution refers to just using executable files provided by the proponent.

53.5 MVC reference software, common conditions, encoder optimizationNo contributions noted (other than AHG input).

54 AVC base specification and related topics

54.1.1.1.1 JVT-W041 ( Prop NN) [A. M. Tourapis, K. Suehring, G. J. Sullivan, A. Leontaris] H.264/MPEG-4 AVC reference software (JM) manual

Revision of the H.264/MPEG-4 AVC Reference Software Manual. JVT decision: Adopt.

198

Further presented on the last day of the meeting. Participants were encouraged to provide further input to improve the software and its associated manual and algorithm description.

54.1.1.1.2 JVT-W042 ( Prop NN) [A. Leontaris, A. M. Tourapis] Rate Control reorganization in the JM reference software

Rate control is an important component of a video compression system as it allows generating compressed bit streams that satisfy bandwidth and buffering constraints. The Joint Model (JM) reference software includes a basic rate control, which, even though not strictly optimal in a rate-distortion sense, allows researchers to evaluate the standard for practical compression scenarios. However, it has been determined that several of the coding tools that are currently included in the JM reference software were not properly supported by the existing rate control algorithm. Other important coding tools, such as hierarchical B-coded pictures, while indirectly supported, were being penalized because the rate control algorithms were never updated to properly consider and take advantage of these tools. On the other hand, the rate control contained severe bugs that were affecting the performance of the software or resulted in invalid bitstreams.

This contribution describes the reorganization of the original rate control algorithm and which was contributed in the Joint Model (JM) 12.0 reference software. This contribution resolved several standing problems that affected the rate control in previous JM versions, but also introduced several new features and support for new tools such as coding of hierarchical structures. More specifically, a number of new rate control modes were introduced to address specific encoding situations, such as intra-only encoding and hierarchical B-coded pictures, without however modifying the essence and basic operation of the original scheme. Instead, the software enhancements have improved the readability and expandability of the original rate control source code, as it was rewritten to adopt an object-oriented structure. The authors note that the presence of broken coding tools in the JM may cause misinterpretation of the actual capabilities of the coding tools.

JVT decision: Adopt.

54.1.1.1.3 JVT-W044 ( Info) [A. M. Tourapis, A. Leontaris, K. Suehring] New JM reference software enhancements

The H.264/MPEG-4 AVC standard has been at times criticized due to its high complexity in terms of both encoding and decoding. Unfortunately, and even though the standard is considerably more complex than previous standards such as H.263 and MPEG2, evaluations on its complexity are sometimes based on the implementation of the AVC Joint Model (JM) reference software. Unlike commercial implementations however, this software was implemented without any complexity considerations. Instead, it was designed mainly with flexibility of implementation in mind since such was required for the proper evolution and development of the standard. The JM codec, and obviously the standard as well, was developed in a relatively significant amount of time and required the involvement of engineers from several companies and institutes with a variety or level of programming knowledge/skills. Although this has helped in the finalization of the standard, its complexity of both the encoder and the decoder has been rather poor compared to almost all commercial or publicly available implementations.

To this purpose, the coordinators of the reference software have undertaken a slow and at times time consuming, effort to reorganize the software, improve its efficiency and coding performance, and at the same time reduce its complexity. One such effort involved the reorganization of most motion compensation and estimation processes within the encoder. This document presents additional enhancements that were introduced to the latest reference software

199

(version 12.2), and which result in considerable complexity reductions at the decoder. The coordinators are still undertaking several other optimizations within the software which may be released in future versions. This contribution would discuss the primary enhancements that were introduced in the JM software.

Some additional needs: Encoder conformance assurance (transform dynamic range, MV area constraint in Baseline), decoder conformance checks (transform dynamic range).

Current “official” version is 12.2.

Observation: At very high bit rates, CAVLC works better than CABAC. Why?

Question: Coordination with VCEG KTA software? Remark: Karsten has broad discretion to coordinate the work on our software – and that presumably includes discretion to coordinate with VCEG efforts.

JVT decision: Adopt.

54.1.1.1.4 JVT-W057 (Late Info) [K. P. Lim] Improved JM text algorithm description

Reference software and descriptions of reference encoding methods and non-normative reference decoding error concealment methods are useful in aiding users of a video coding standard to establish and test conformance and interoperability, and to educate users and demonstrate the capabilities of the standardThis document specifies non-normative reference encoding methods and methods of concealing errors and losses in decoders for video data conforming to ITU-T Recommendation H.264 | ISO/IEC International Standard ISO/IEC 14496-10 advanced video coding.

JVT decision: Adopt.

54.1.1.1.5 JVT-W140 -B (BoG) [T. Suzuki] Toward a professional profiles conformance specification

Report of preliminary work toward a professional profiles conformance specification. JVT decision: Endorsed.

55 Video annotation (jointly discussed with MPEG requirements 3:30 pm Wednesday 25 April)

55.1.1.1.1 JVT-W032 ( Info) [Q. Chen, C. Louis, Z. Chen] Requirements of video annotation in video coding

This document presents some asserted requirements on adding video annotation support into a video coding standard. Some application scenarios were listed which can reportedly benefit from this practice, and these were placed into three categories: text annotation, visual characteristics, and video structure. The contribution recommended that the JVT work out some methods to support the target applications.

Presentation and discussion were held together for JVT-W032, JVT-W033, and JVT-W034.

200

Asserts that MP4 file format metadata tracks and MPEG EPG tracks do not fully address current needs.

Suggests supporting carriage of annotation data in both the system level and video bitstream level. Support within video is asserted to be useful due to the ability to carry the metadata regardless of the system environment.

Asserts that MPEG-7 has too many things in it – to the degree that people don’t know which ones to use. Potential approaches to this asserted problem include profiles of MPEG-7 or non-normative guidance about which MPEG-7 data types to use.

Remark: Putting such data into the video layer means that you need to touch the video layer just to manipulate the metadata, and that you may need to search through very high bit rate information to locate metadata of interest.

Remark: How does this connect with compression work?

It was suggested that metadata should be based on MPEG-7 as much as possible, and that duplications of effort and text and inconsistencies of design should be avoided. Work should be kept coordinated across the organizational boundaries.

Where to carry? Systems layer or video layer (SEI)? New metadata types? Re-use MPEG-7? With modified structure?

Arguments for doing it in video: Persistent regardless of type of systems and FF, can be generated as part of encoding, good to have it as part of raw video stream.

Elaborate pro’s and con’s doing it here or there. It may also be the case that for certain cases one or the other is better. Even then, the metadata should be compatible (same subset of MPEG-7 etc.).

Explore relation between metadata and coding. Metadata should be MPEG-7.

AHG to be established in MPEG on the topic.

55.1.1.1.2 JVT-W033 ( Prop 2.2/3.1) [Q. Chen, Z. Chen, X. Gu] Video annotation SEI message

This document proposes to add video annotation SEI messages into AVC bit streams to add capabilities for video searching, browsing, and other applications. A couple of related issues are discussed and finally a particular approach is proposed to the JVT.

See notes relating to JVT-W032.

55.1.1.1.3 JVT-W034 ( Prop 2.2/3.1) [C. Louis, O. Lionel, L. Frederic, Z. Chen, Q. Chen] Fingerprint and video structure for video annotation SEI message

This document proposes to add “video fingerprint” and video structure support in SEI messages for video annotation. These are combined into a proposed video annotation SEI message. The applications are reported to be fast video copy detection, fast video browsing, etc.

See notes relating to JVT-W032.

201

56 AVC errata and clarification issues

56.1.1.1.1 JVT-W134 -Q (Late Prop 2.2) [S. Narasimhan] Splicing issues and some suggested changes

This contribution was subject to lateness penalties as recorded elsewhere in this report.

Splicing is currently used in U.S. cable networks for digital ad-insertion based on the MPEG-2 video standard and there are plans to migrate these applications based on AVC in the near future. In these applications, the splicing equipment (or function) combines two independently encoded AVC streams and is expected to produce an AVC ‘conformant’ output for receiving equipment. This contribution outlines issues related to generating an AVC conformant output by such splicing equipment and suggests some reportedly-minor changes to the AVC standard to reportedly assist these applications.

Remark: What about SMPTE RP 312M on seamless splicing? Basically, that is not being used, and is reportedly expected be withdrawn.

ITU-T J.181 is relevant (developed by SCTE and brought to ITU-T SG 9 or 11 with further involvement by Japan ITU-T members).

Discusses local ad insertion, other types of splicing, and associated difficulties.

Output document JVT-W210 to be produced incorporating issues noted herein and others identified by the editor of the output document, Gary J. Sullivan.

57 Requirements joint discussions with WG 11

Joint discussions were held with WG 11 requirements and video subgroups at 2pm on Wednesday 25 April. Some issues raised in WG 11 documents were discussed.

Also see notes relating to video annotation and profiles and applications.

57.1.1.1.1 M14452 WG 11 input [T. Murakami, K. Asai, Y. Yamada] Requirement of full-color video coding for consumer applications

This WG 11 contribution suggested considering consumer-device support for the following.– 4:4:4 chroma– More than 8 bit dynamic range

Picture formats from QVGA to HDTV.

Request for 4:4:4 in consumer applications. Camcorders and big displays would reportedly benefit. Real big formats (8K) expected to happen after 2010. Requirements: frame rate up to 60 Hz, progressive scanning, up to 10 bits, enhanced coding efficiency for 4:4:4. Comment: Could be achieved with 4:4:4 predictive profile. What would be needed that is different than that? Are bit rate ranges of the proposal useful? Perform a study about exact requirements and identify whether the professional profiles available would require any change.

Remark: Below certain bit rates/fidelities, having 4:4:4 and high bit depth hurts rather than helps (some skepticism about that remark was expressed). Remark: There were old contributions from Dolby and recent contributions in VCEG that may help clarify these issues.

202

Having a better understanding of those issues is needed, and an understanding of what is needed that would be different than what the High 4:4:4 Predictive profile provides.

57.1.1.1.2 M14360 [USNB to WG 11] Issues relating to expiring patentsThe USNB to WG 11 noted 1) That some number of core patents in media coding have expired or will be expiring soon; 2) That there also exist un-patented technologies in media coding; 3) That for many years the combination of CPU power, bandwidth, and compression efficiency was not sufficient to give acceptable quality in many environments, and improved compression efficiency was the driving factor in developing new standards, but for at least some environments this has now changed – indeed, the USNB asserts that there are striking examples where not all the CPU power available is used, or not all the bandwidth is used; there are also environments where the strongest compression is not the dominant selection criterion; 4) That it has been argued that a royalty-free standard would detrimentally affect the uptake of existing MPEG standards – however, if it is technically possible to develop a standard which does this, the USNB prefers that it be done in WG 11 where there is expertise in doing it well, and where such a putative standard could be made a 'family member' with other MPEG standards (with an upgrade path, for example, or related technical ‘roots’ etc.); 5) That the 'terms of engagement' of a study on developing a process for royalty-free standards, and the results and follow-on for such work, should be made more clear before more discussion is held at WG 11.

M14360: “No explicit request. Main purpose to create a discussion. No request to take specific action.” Question raised: Does WG 11 have the expertise to find out whether a standard is royalty-free?

No explicit request for action was made in these comments, and it was asserted that there would need to be a more clear understanding of a process for developing royalty-free standards prior to proceeding with such work.

57.1.1.1.3 JVT-W127 ( Req) [M. Tanimoto, T. Fujii, H. Kimata, S. Sakazawa] Requirements for FTV (MPEG M14417)

Proposed requirements for FTV (free viewpoint television) are provided in this document. The content of this document is the same as in the MPEG document M14417.

FTV is anticipated to be a “ray-based system” as opposed to the “pixel-based system” approach largely taken in today’s video coding standards. FTV should be able to be viewed on a wide variety of displays, including both 2D and 3D displays, on a wide variety of platforms (from mobile phones to fixed large room-based displays). Considers need for view generation, including depth map determination and interpolation for display.

Note relevance of ISO/IEC 23002-3 and JVT-W100.

Proposed standardization action items for FTV1) FTV data format2) Compression3) Rendering4) Transmission data format and protocol (ITU-T SG 9 working on this)

There was some discussion of the distinction between “data format” and “compression”.

203

There was some discussion of the scope of standardization – and particularly regarding whether “rendering” should be standardized, and where the compression decoding process ends and rendering begins.

FTV must support many types of displays. Function of view generation should be simple. FTV requires depth search and interpolation. 3 possibilities: Both at sender (ray space), both at receiver (MVC plus postprocessing), or separated (search at sender, interpolation at receiver) (MVC plus depth). Supported format is view plus depth, but information about reflection also might need to be included in FTV data. Standardisation issues: Format, compression, rendering, transmission. Proposal: Determine FTV format together with rendering; compression format can be extension of MVC.

Data format would consist of all the information needed to perform a good rendering (image, depth, illumination). Compression for most compact representation of data with a certain quality.

Question: To which extent should the method of creating additional views be specified? Unclear – rendering would definitely be needed to be standardized when it is used as part of the (de-)compression. Otherwise, how can it be known which renderer is required which is display specific?

Concern (against normative rendering) that freedom in the design of display is given up (in terms of quality, complexity, …). However, testing with rendering (maybe different) will be needed.

See also related notes in section on JVT-W100.

Fernando will draft an MPEG requirements document on application requirements of FTV.

Can very well be seen as extension of current MVC. However, exploiting relationships between image, depth and reflection could lead to better compression than separate handling.

Changes relative to prior MPEG MVC Requirements document have been agreed:– Spatial scalability as “shall”– Variation of spatial and temporal resolution across views as “shall”

58 JVT internal operating rulesJVT decision: The following clarifications/adjustments of JVT operating rules have been adopted.

The JVT decided that participants shall to refrain from long (=more than 4 Minutes) presentations of their proposal, if the results of their coding efficiency experiments have provided less than 2% bit-rate on average (or equivalently 0.1 dB gain on average).

Presentations should also not use "cherry picking" of results for summary reporting in abstracts and presentations. Summary reports must be true summaries – not highlights of best results while ignoring worst results.

Regarding late contributions: Due to our difficulties with a large quantity of late-submitted contributions at this and other recent meetings, the JVT has agreed that for its next meeting, no late-uploaded (non-AHG-report, non-liaison, non-verification) contribution will be presented without having a minimum of 4 JVT participants (working for organizations other than that of the primary contribution author) recorded by name as supporting the allowance of such a presentation, in addition to a consensus of the general JVT membership to allow the presentation.

204

Such support to allow a presentation is to be understood to not necessarily imply support of the adoption of the content of the late contribution, but only as a positive expression that the document should be allowed to be presented. Additionally, the provider of such a presented late contribution shall send an email apology to the JVT email reflector. This rule does not apply to material requested by the JVT at the meeting (e.g., reports of JVT-authorized side activities).

For all contributions that have presentation material that is used to present them to the group (e.g., PowerPoint presentations), the presentation material should be provided along with the written contribution (within the same zip container file). PDF is preferred over PPT for presentations when the PPT filesize is large and there is no need for the slide deck to be editable by others.

All submissions must be made in JVT-Wxxx.zip format with the word docs, excel sheets and other information being in the zip container. The document must contain an abstract and be accompanied with an e-mail notification containing title, authors and abstract (identical to the one in the doc) which is no longer than 200 words and is written in 3rd person in a manner that does not express endorsement of the content of the document.

On filenames inside of .zip containers – use a filename so that if you take the files out of the zip container, you'll still know what contribution they came from. Every file in the .zip container for document JVT-Wxxx should start with JVT-Wxxx. Example: JVT-Wxxx.doc (main document), JVT-Wxxx_presentation.pdf, JVT-Wxxx_results1.xls, etc.

When providing additional or revised files, do not include copies of files that were already included in the prior .zip archive for the same contribution and do not re-use the same filenames without adding revision numbers (r1, r2, etc.) – this saves us needing to worry about whether the files we get with the same filenames are the same or different.

Independent verification (necessary for adoption of a proposal) is provided either througha) independent implementation by 1 or more company different than the proponent based on

the textual description (after adoption, both decoder source code versions must be made publicly available and one encoder version), or

b) providing source code to all CE participants prior to the meeting (CEs can only be joined at the meeting, when the CE is created. CEs are created at each meeting and last until the next meeting.)

Simply running binary executables provided by a proponent is not ordinarily considered independent verification. Source code should be provided and used, and the verifying party should invest a proper degree of effort to ensure that the “verification” they perform is a meaningful and professional study with significant depth rather than just a perfunctory procedural formality.

For every SEI message and every syntax element that are currently in the SVC/MVC draft, a showcase has to be provided in order to retain it in the JSVM/JMVM/JD. If such a showcase is not provided at the next meeting for an SEI message or parts of it, the SEI message or the respective parts will be removed from the JSVM/JMVM/JD. The source code and executables for the showcase must be made available.

A first CE description should be available at the last day of the meeting. Changes of the CE description are only allowed until 3 weeks prior to the next meeting. These changes must be of evolutionary characteristic relative to the input documents on which the CE is based and must be agreed by those who contributed the respective input document(s) or be added as an option.

205

Contributions that are proposals of new technology that was not what was described as being tested in a CE (even if related to the tested technology) should not indicate that they are CE documents in their title and abstract.

59 List of adoptionsThis section of the report lists adoption actions by the JVT at this meeting in condensed form. All items noted in this section should be redundant with actions noted elsewhere in this report. Where listed, the person listed in brackets is responsible for provision of text and software integration.

59.1 SVC normative adoptions into JDAdoption actions are listed as follows:– FGS: JVT-W090.– IntraBL treated as inter for constrained intra pred (see notes on JVT-W090).– Intra MBs in base layer not exceeding IntraBL by more than 1.5 (see notes on JVT-W090).– ESS improvement: JVT-W030.– Interlaced restrictions: JVT-W025.– Remove SRP (see notes on JVT-W026 and JVT-W118).– De-blocking JVT-W063r1.– Header re-writing JVT-W046.– Inheritance of deblocking control (see notes on JVT-W046).– Pictures not for output JVT-W047.– Various items JVT-W048.– Profile & bit rate indicators (subset of JVT-W051).– Integrity check JVT-W052.– MBs required for picture only for QID = 0 (see notes JVT-W052).– Quality layer SEI syntax JVT-W137.– Priority ID JVT-W053r2.– Seven restriction indicators in scalability info SEI (see notes on JVT-W064).– Various HL syntax issues (see notes on JVT-W125).– Redundant pictures into profile A & SEI messages (JVT-W049).– SEI message tl0_pic_idx (sec. 3.3 of JVT-W062r3).– Profile changes as recorded in profiles section.– Change to scaling in position calc for large pictures (see notes on JVT-W136).

59.2 SVC normative adoptions into JSVMAdoption actions are listed as follows:– FGS modifications JVT-W119.– FGS modifications JVT-W121r1 (which combines elements of JVT-W111 and JVT-W121).– Dyadic subband coding method JVT-W097.

59.3 SVC non-normative adoptionsAdoption actions are listed as follows:– Encoder problem detection trick from JVT-W105.– Rate control JVT-W043

206

59.4 SVC software adoptionsNo particular adoption actions noted.

59.5 MVC normative JD adoptionsAdoption actions are listed as follows:– Various high-level syntax changes JVT-W035.– Signal views to be output JVT-W036.– View scalable SEI JVT-W037.– nal_ref_idc_view JVT-W056.– Reference picture list reordering bug fix JVT-W066.– Parallel decoding SEI syntax modifications (as presented in JVT-W080).– Camera parameters & max disparity JVT-W060.– Restriction of temporal direct and weighted prediction (see notes on JVT-W040).

59.6 MVC JMVM adoptionsAdoption actions are listed as follows:– Deblocking filter control JVT-W024.– Illumination compensation info derivation JVT-W031.– MVC Motion skip mode JVT-W081 as recorded in JVT-W139.

59.7 MVC non-normative adoptionsNo particular adoption actions noted.

59.8 JM non-normative adoptionsAdoption actions are listed as follows:– JM manual JVT-W041– JM rate control JVT-W042– JM software cleanup JVT-W044– JM algorithm text description JVT-W057

59.9 Other normative adoptionsNo particular adoption actions noted.

59.10 Other non-normative adoptionsNo particular adoption actions noted.

60 Software integration planDelegated to the software coordinators.

61 SVC conformance work planThe Hangzhou meeting report recorded the following: “The following companies each announce to provide at least 10 conformance bitstreams for SVC: HHI, Sharp, Thomson, RWTH (maybe), Nokia (potentially), Orange, Microsoft, Qualcomm.”

207

These parties were not present on Tuesday morning. It was asserted that a conformance workplan working draft (WD) needed urgently to be set up by end of week. Progress was later reported from a break-out group activity as recorded in JVT-W138.

62 SVC verification test plan

Action items noted during the meeting:– Viewing of available material– Clarify situation about the completeness of the test material for SVC compression

performance– demonstrating potential prototype applications of SVC (e.g. showing advantage of scalability

in streaming)

Report of breakout work was presented:– Review of JVT-W131: Bitrate SNR may be a bit too high; ratio 3:1 Enh:Bas better than 2:1,

eventually include up to HD– Bandwidth fluctuation scenario may eventually not be too useful without FGS (or would

need implementation of concealment which might be difficult)– Profile B: Broadcast SD & HD– Scalability between 1080i and 1080p might also be a convincing scenario

63 List of AHGs establishedThe following JVT “ad hoc groups” (AHGs) were established to progress work on identified topics until the next meeting of the JVT.

63.1 JVT project management and errata reportingDiscussion: jvt-experts@lists.rwth-aachen.deChair: Gary Sullivan, Jens Rainer Ohm, Ajay Luthra, and Thomas WiegandMandates:– Collect errata reports on standards under management of JVT– Coordinate overall interim JVT progress– Prepare status information for JVT status reporting

63.2 JM Text, reference software, bitstream exchange and conformanceDiscussion: jvt-experts@lists.rwth-aachen.deChair: Thomas Wiegand, Karsten Sühring, Alexis Tourapis, Teruhiko Suzuki, Keng Pang LimMandates:– Maintain and update JM algorithm description text– Maintain and update JM reference software and its usage manual– Facilitate exchange of test bitstreams to aid interoperability testing– Collect bitstreams for inclusion in Conformance specifications– Identify and correct problems in Conformance specifications and associated bitstreams

63.3 AVC professional applicationsDiscussion: jvt-experts@lists.rwth-aachen.deChair: Teruhiko SuzukiMandates:– Finalize software for new professional profiles– Collect bitstreams for Conformance specification update for new prof profiles

208

63.4 SVC JD and JSVM text, software and conformanceDiscussion: jvt-svc@lists.rwth-aachen.deChair: Heiko Schwarz, Jérome Vieron, Thomas Wiegand, Mathias Wien, Alex Eleftheriadis, Vincent BottreauMandates:– Edit and deliver JD and JSVM text– Coordinate JSVM software integration– Coordinate bug-fixing process for the JSVM software– Maintain JSVM software manual– Plan, edit, and collect bitstreams for SVC conformance specification

63.5 SVC bit depth and chroma format scalabilityDiscussion: jvt-svc@lists.rwth-aachen.deChair: Yongying Gao, Andrew Segall, Thomas WiegandMandates:– Identify applications– Work out suggestions for detailed needs– Find/create test material– Study bit-depth reduction techniques, e.g., tone-mapping tools– Study color space and/or gamma conversion requirements– Study combined spatial and bit depth scalability– Define experiments and test conditions– Investigate software and text modification needs– Identify complexity issues

63.6 SVC FGS applications and design simplificationDiscussion: jvt svc@lists.rwth-aachen.deChair: Justin Ridge, Marta KarczewiczMandates:– Identify applications for FGS and their characteristics– Define experiments and test conditions relating to FGS technology– Explore simplification of FGS design

63.7 MVC high-level syntax and buffer managementDiscussion: jvt-mvc@lists.rwth-aachen.deChair: Anthony Vetro, Purvin PanditMandates:– Discuss high-level syntax for MVC including NAL unit type, NAL unit header extension,

SPS extensions, slice layer and integration with SVC syntax.– Discuss reference picture management to enable simultaneous picture output of different

views and to facilitate parallel processing.– Discuss issues related to HRD.– Propose refined syntax and decoding processes for JMVM.

209

63.8 MVC JD and JSVM text and softwareDiscussion: jvt-mvc@lists.rwth-aachen.deChair: Hideaki Kimata, Aljoscha Smolic, Purvin Pandit, Anthony Vetro, Chen YingMandates:– Collect comments on draft, perform necessary editing and delivery.– Maintain JMVM and JD document and collect comments on the text.– Coordinate JMVM software integration– Coordinate bug-fixing process for the JMVM software– Maintain JMVM software manual

63.9 MVC experimental framework and testing conditionsDiscussion: jvt-mvc@lists.rwth-aachen.deChair: Hideaki Kimata, Aljoscha SmolicMandates:– Evaluate application needs in MVC framework– Discuss testing conditions to evaluate specific application needs– Consider needs for new tools to be evaluated

63.10 MVC solutions using existing AVC decodersDiscussion: jvt-mvc@lists.rwth-aachen.deChair: Purvin PanditMandates:– Collect comments on methods for enabling AVC decoding of multiview video

(spatial/temporal/others)– Study the complexity of such methods– Investigate the applications enabled

63.11 MVC reduced resolution update, downsampled reference and adaptive reference filtering

Discussion: jvt-mvc@lists.rwth-aachen.deChair: Purvin Pandit, Hideaki Kimata– Investigate approaches for enhancing MVC coding efficiency using spatial downsampling– Evaluate the complexity of such methods– Investigate the relationship between downsampling approaches and view interpolation– Evaluate subjective quality associated with methods– Study the complexity associated with adaptive reference filtering– Evaluate performance of adaptive reference filtering under JMVM common conditions

64 Resolutions conveyed to MPEG parent bodyThe JVT approved the following resolutions for conveyance to its MPEG (WG 11) parent body.

64.1 Resolutions relating to ISO/IEC 14496-464.1.1 The JVT and the video subgroup recommend to approve the following documents

No. Title TBP Available14496-4 Conformance testing

8954 Request for ISO/IEC 14496-4:2004/Amd.30 No 07/04/27210

8955JVT-W204

Working Draft 1 of ISO/IEC 14496-4:2004/Amd.30 Conformance testing for new profiles for professional applications

No 07/06/29

8956 Request for ISO/IEC 14496-4:2004/Amd.31 No 07/04/278957JVT-W205

Working Draft 1 of ISO/IEC 14496-4:2004/Amd.31 Conformance testing for SVC profiles

No 07/06/29

64.1.2 The JVT and the video subgroup thank the following companies for their commitment to provide conformance testing streams for ISO/IEC 14496-4:2004/Amd.30: Mitsubishi Electric Corp., Panasonic, Sejong University, Thomson.

64.1.3 The JVT and the video subgroup thank the following companies for their commitment to provide conformance testing streams for ISO/IEC 14496-4:2004/Amd.31: ETRI, FhG-HHI, France Telecom/Orange, Layered Media, Sharp, Thomson.

64.2 Resolutions relating to ISO/IEC 14496-564.2.1 The JVT and the video subgroup recommend to approve the following documents

No. Title TBP Available14496-5 Reference Software

8958 Request for ISO/IEC 14496-5:2001/Amd.18 No 07/04/278959JVT-W206

Working Draft 1 of ISO/IEC 14496-5:2001/Amd.18 Reference software for new profiles for professional applications

No 07/06/29

8960 Request for ISO/IEC 14496-5:2001/Amd.19 No 07/04/278961JVT-W211

Working Draft 1 of ISO/IEC 14496-4:2001/Amd.19 Reference software for SVC

No 07/06/29

64.3 Resolutions relating to ISO/IEC 14496-1064.3.1 The JVT and the video subgroup recommend to approve the following documents

No. Title TBP Available14496-10 Advanced Video Coding

8962JVT-W201

Study Text (version 3) of ISO/IEC 14496-10:2005/FPDAM3 Scalable video coding

No 07/05/31

8963JVT-W202

Joint scalable video model (JSVM) 10 No 07/05/31

8964JVT-W203

JSVM 10 software No 07/06/29

8965JVT-W212

Draft SVC verification test plan version 3.0 No 07/05/18

8966JVT-W209

Working Draft 3 of ISO/IEC 14496-10:2005/Amd.4 Multiview video coding

No 07/05/18

8967JVT-W207

Joint multiview video model (JMVM) 4 No 07/05/18

8968JVT-W208

JMVM 4 software No 07/05/31

64.3.2 The JVT and the video subgroup request WG 11 National Bodies to kindly consider the SVC Study Document N8962 [JVT-W201] and if necessary provide additional comments by the July 2007 meeting.

211

64.4 Resolutions relating to future meeting scheduling64.4.1 The JVT chairmen propose to hold a JVT meeting during June 29 through July 6,

2007 under the auspices of the meeting of ITU-T SG 16 in Geneva, CH. Further meetings are proposed to be held during October 19-26, 2007 under WG 11 auspices in Shenzhen, CN, and during January 11-18, 2008 under WG 11 auspices in Antalya, TR.

64.5 Resolutions relating to ad hoc group activities64.5.1 The JVT provides the following list of JVT ad hoc groups appointed to progress

work in the interim period until the next JVT meeting:

Title and Email Reflector Chairs MtgJVT project management and errata reporting(jvt-experts@lists.rwth-aachen.de)

Gary Sullivan, Jens Rainer Ohm, Ajay Luthra, and Thomas Wiegand

N

JM Text, reference software, bitstream exchange and conformance(jvt-experts@lists.rwth-aachen.de)

Thomas Wiegand, Karsten Sühring, Alexis Tourapis, Teruhiko Suzuki, Keng Pang Lim

N

AVC professional applications(jvt-experts@lists.rwth-aachen.de)

Teruhiko Suzuki N

SVC JD and JSVM text, software and conformance(jvt-svc@lists.rwth-aachen.de)

Heiko Schwarz, Jérome Vieron, Thomas Wiegand, Mathias Wien, Alex Eleftheriadis, Vincent Bottreau

N

SVC bit depth and chroma format scalability (jvt-svc@lists.rwth-aachen.de)

Yongying Gao, Andrew Segall, Thomas Wiegand

N

SVC FGS applications and design simplification(jvt-svc@lists.rwth-aachen.de)

Justin Ridge, Marta Karczewicz N

MVC high-level syntax and buffer management(jvt-mvc@lists.rwth-aachen.de)

Anthony Vetro, Purvin Pandit N

MVC JD and JMVM text and software(jvt-mvc@lists.rwth-aachen.de)

Hideaki Kimata, Aljoscha Smolic, Purvin Pandit, Anthony Vetro, Chen Ying

N

MVC experimental framework and testing conditions(jvt-mvc@lists.rwth-aachen.de)

Hideaki Kimata, Aljoscha Smolic N

MVC solutions using existing AVC decoders(jvt-mvc@lists.rwth-aachen.de)

Purvin Pandit N

MVC reduced resolution update, downsampled reference and adaptive reference filtering(jvt-mvc@lists.rwth-aachen.de)

Purvin Pandit, Hideaki Kimata N

65 AttendancePersons registered to attend the meeting, as recorded by a sign-in sheet circulated during the meeting, were the following (185 listed participants):1) Alvarez, José Roberto (Mobilygen)2) Amon, Peter (Siemens AG)3) Bandoh, Yukihiro (NTT)4) Bao, Yiliang (Qualcomm)5) Baik, Sung Uk (Oniontech)

212

6) Bivolarski, Lazar (Brightscale)7) Bjøntegaard, Gisle (Tandberg)8) Borgwardt, Peter (Motorola)9) Bottreau, Vincent (Thomson R&D France)10) Bourge, Arnaud (Philips / NXP)11) Branguolo, Sebastien (SSM)12) Bruls, Fons (Philips)13) Cammas, Nathalie (Orange - France Telecom.)14) Chen, Lulin (Omneon Video Networks USA)15) Chen, Quqing (Thomson)16) Chen, Weizhong (Huawei Tech.)17) Chen, Ying (Tampere Univ. Tech.)18) Cheong, Hye-Yeon (Univ. Southern California)19) Chiu, Yi-Jen (Intel)20) Choi, Byeongho (KETI)21) Choi, Hae-Chul (ETRI)22) Choi, Jongbum (Samsung)23) Choi, Woongil (Samsung AIT)24) Chujoh, Takeshi (Toshiba)25) Chung, Hyukjune (Qualcomm)26) Cieplinski, Leszek (Mitsubishi Electric)27) Civanlar, M. Reha (DoCoMo Labs USA)28) Cock, Jan De (Ghent Univ.)29) Cornog, Katie (Avid)30) Coté, Guy (Mobilygen)31) Divorra, Òscar (Thomson)32) Eleftheriadis, Alex (Layered Media)33) Fröjdh, Per (Ericsson)34) Fujii, Toshiaki (Nagoya Univ.)35) Gao, Yongying (Thomson)36) Gallant, Michael (LSI Logic, Canada)37) Goh, Kwong Hueng (Inst. for Infocomm Research)38) Guleryuz, Onur (Docomo USA Labs)39) Han, Woo-Jin (Samsung)40) Hannuksela, Miska (Nokia)41) Harmani, Oztan (DoCoMo USA Labs)42) Haskell, Barry (Apple)43) He, Jones (Freescale)44) Hinds, Arianne (IBM)45) Ho, Yo-Sung (GIST)46) Hong, Danny (Layered Media)47) Hsiang, Shih-Ta (Motorola)48) Huang, Wei-Hung (MediaTek)49) Huang, Yu-Wen (MediaTek)50) Huo, Junyan (Xidian Univ.)51) Ishtiaq, Faisal (Motorola)52) Itoh, Takashi (Fujitsu Labs)53) Jeon, Byeong-Moon (LG Electronics)54) Jeon, Byeungwoo (SKKU)55) Jeon, Yongjoon (LG Electronics)56) Jia, Jie (Sejong Univ.)57) Jung, Bongsoo (SKKU)58) Jung, Joël (France Telecom R&D)

213

59) Kang, Jung Won (ETRI)60) Kanumuri, Sandeep (NTT DoCoMo USA Labs)61) Karczewicz, Marta (Qualcomm)62) Kim, Dongkyun (Sejong Univ.)63) Kim, Hae Kwang (Sejong Univ.)64) Kim, Hyun Mun (Samsung AIT)65) Kim, Jae Hoon (Univ. Southern California)66) Kim, Jinwoong (ETRI)67) Kim, Jong Lak (DSP Group)68) Kim, So Young (Samsung Electronics)69) Kim, Yong-Hwan (KETI)70) Kimata, Hideaki (NTT)71) Kimoto, Takahiro (NEC)72) Koo, Han-Suh (LG Electronics)73) Kopansky, Arkady (Sarnoff)74) Lainema, Jani (Nokia)75) Lee, Sang-Heon (Seoul Natl. Univ.)76) Lee, Sang-Houn (DSP Group)77) Lee, Yung Ki (Sejong Univ.)78) Lee, Yung-Lyul (Sejong Univ.)79) Lei, Shawmin (Sharp Labs USA --> MediaTek)80) Leontaris, Athanasios (Dolby)81) Li, Zhengguo (I2R)82) Lim, Chong Soon (Panasonic)83) Lim, Sung Chang (Sejong Univ.)84) Lin, Sixin (Huawei)85) Lu, Ning (Intel)86) Luo, Jiancong (Thomson)87) Luthra, Ajay (Motorola)88) Masashi, Takahashi (Hitachi)89) Matsubara, Akio (Ricoh)90) McCartley, Sean (Modulus Video)91) Meany, James (Boeing)92) Müller, Karsten (Fraunhofer HHI)93) Naito, Sei (KDDI)94) Nakamura, Hiroya (JVC)95) Narasimhan, Sam (Motorola)96) Ndili, Obianuju (Santa Clara Univ.)97) Nilsson, Mike (BT)98) Nishi, Takashi (Oki Electric Industry)99) Ogunfunmi, Tokunbo (Santa Clara Univ.)100) Oh, Kwan-Jung (GIST)101) Ohm, Jens-Rainer (RWTH Aachen Univ.)102) Onno, Patrice (Canon France)103) Pandit, Purvin (Thomson)104) Park, Ji Ho (KETI)105) Park, Min-woo (Kyung Hee Univ.)106) Park, Seanae (Kwangwoon Univ.)107) Park, Seung-Wook (LG Electronics)108) Pateux, Stephane (Orange - France Telecom)109) Peng, Wen Hsiao (Samsung AIT)110) Pereira, Fernando (IST)111) Prieto, Yolanda (Freescale)

214

112) Ransburg, Michael (Klagenfurt Univ.)113) Rathgen, Thomas (Ilmenau Univ.)114) Regunathan, Shankar (Microsoft)115) Reznik, Yuriy (Qualcomm)116) Ridge, Justin (Nokia)117) Rault, Patrick (Quartics)118) Rodriguez, Arturo (Scientific Atlanta / Cisco)119) Sakazume, Satoru (JVC)120) Sampedro, Jesus (Polycom)121) Sato, Kazushi (Sony)122) Schwarz, Heiko (Fraunhofer HHI)123) Schierl, Thomas (Fraunhofer HHI)124) Segall, Andrew (Sharp Labs USA)125) Sekiguchi, Shun-ichi (Mitsubishi)126) Senoh, Takanori (Univ. Tokyo)127) Seo, Chang-Won (Sejong Univ.)128) Seo, Juheon (Sejong Univ.)129) Seo, Jungdong (Yonsei Univ.)130) Shi, Xiaojin (Apple)131) Shim, Woo-Sung (Samsung Electronics)132) Shimizu, Shinya (NTT)133) Shiodera, Taichiro (Toshiba)134) Sim, Donggyu (Kwangwoon Univ.)135) Sjöberg, Rickard (Ericsson)136) Smolić, Aljoscha (Fraunhofer HHI)137) Su, Yeping (Thomson USA --> Sharp USA)138) Suh, Doug Young (KHU)139) Suh, Jong-Yeul (LG Electronics)140) Sullivan, Gary (Microsoft Corp.)141) Sun, Huifang (Mitsubishi)142) Suzuki, Teruhiko (Sony)143) Takamura, Seishi (NTT)144) Tam, James (CRC, Canada)145) Tan, Thiow Keng (NTT DoCoMo)146) Tanimoto, Masayuki (Nagoya Univ.)147) Tanizawa, Akiyuki (Toshiba)148) Thoma, Herbert (Fraunhofer IIS)149) Tian, Dong (Thomson)150) Timmerer, Christian (Klagenfurt Univ.)151) Topiwala, Pankaj (FastVDO)152) Tourapis, Alexandros (Dolby Labs)153) Tung, Yi-Shin (Setabox Tech. Corp.)154) Ugur, Kemal (Nokia)155) Van de Walle, Rik (Ghent Univ.)156) Vetro, Anthony (Mitsubishi Electric)157) Vieron, Jerome (Thomson R&D France)158) Viscito, Eric (eV Consulting)159) Wan, Wade (Broadcom)160) Wang, Haohong (Marwell)161) Wang, Xianglin (Nokia)162) Wang, Yong (Motorola)163) Watanabe, Hitoshi (Qpixel)164) Wedi, Thomas (Panasonic)

215

165) Wiegand, Thomas (Fraunhofer HHI)166) Wien, Mathias (RWTH Aachen Univ.)167) Wittmann, Steffen (Panasonic)168) Wu, Ping (Tandberg Television)169) Wus, John (Panasonic)170) Xiong, Lianhuan (Huawei)171) Xu, Xiaozhong (Tsinghua Univ.)172) Yagasaki, Yoichi (Sony)173) Yamamoto, Tomoyuki (Sharp)174) Yamasaki, Takahiro (Oki Electric Industry)175) Yang, Haitao (Xidian Univ.)176) Yang, Jeong-Hyu (LG Electronics)177) Yang, Jungyoup (SKKU)178) Yang, Ping (Tsinghua Univ.)179) Yao, Wei (I2R)180) Ye, Yan (Qualcomm)181) Yoo, Jeong-Ju (ETRI)182) Yu, Haoping (Thomson)183) Yu, Lu (Zhejiang Univ.)184) Zhang, Liang (CRC, Canada)185) Zheng, Jianhua (Huawei)

216

Annex J – Audio report

Source: Schuyler Quackenbush, Chair, Audio Subgroup

1 Opening of the meeting.........................................................................................................2162 Administrative matters...........................................................................................................216

2.1 Approval of previous meeting report 2162.2 Approval of agenda and allocation of contributions 2162.3 Task Groups 2162.4 Communications from the Chair 2162.5 Joint meetings 2162.6 Received National Body Comments and Liaison matters 216

3 Record of AhG meetings.......................................................................................................2164 Audio plenary, joint meeting and task group activities.........................................................216

4.1 Review of AHG reports 2174.2 Received national body comments and liaison matters 2174.3 Joint Meetings217

4.3.1 Systems at Audio on MP4FF and Sampling Rate.......................................................2174.4 Task Group discussions 217

4.4.1 MPEG Surround..........................................................................................................2174.4.2 SAOC...........................................................................................................................2184.4.3 MPEG-4 ELD..............................................................................................................2194.4.4 Speech and Audio Exploration....................................................................................2234.4.5 Symbolic Symbol Representation................................................................................2234.4.6 MPEG-1, MPEG-2 and MPEG-4 audio, conformance, reference software................224

5 Meeting deliverables..............................................................................................................2255.1 Recommendations for final plenary 2255.2 Establishment of Ad-hoc Groups 2255.3 Approval of output documents 2255.4 Responses to Liaison and NB comments 2255.5 Press statement 225

6 Future activities.....................................................................................................................2256.1 Schedule of future meetings 2256.2 Agenda for next meeting 2256.3 All other business 2256.4 Closing of the meeting 225

Annex A Participants...............................................................................................................226Annex B Audio Contributions and Schedule...........................................................................227Annex C Task Groups.............................................................................................................232Annex D Output Documents....................................................................................................233Annex E Agenda for the 81st MPEG Audio Meeting.............................................................235

217

1 Opening of the meetingThe MPEG Audio Subgroup meeting was held during the 80th meeting of WG11, April 23-27, 2007, San Jose, CA, USA. The list of participants is given in Annex A.

66 Administrative matters66.1 Approval of previous meeting reportThe 79th Audio Subgroup meeting report was registered as a contribution, and was approved.

66.2 Approval of agenda and allocation of contributionsThe agenda and schedule for the meeting was discussed, edited and approved. It shows the documents contributed to this meeting and presented to the Audio Subgroup, either in the task groups or in Audio plenary. The Chair brought relevant documents from Requirements, Systems and MDS to the attention of the group. It was revised in the course of the week to reflect the progress of the meeting, and the final version is shown in Annex B.

66.3 Task GroupsTask groups were convened for the duration of the MPEG meeting, as shown in Annex C. Results of task group activities are reported below.

66.4 Communications from the ChairThe Chair summarised the issues raised at the Sunday evening Chair’s meeting, proposed task groups for the week, and proposed agenda items for discussion in Audio plenary.

66.5 Joint meetingsThe joint meetings with Audio over the course of the week are listed here and are reported on below.Groups What Where Day TimeSystems, Audio

14529, MP4 file format considerations for high sample-rate audio

Audio Wed 1130-1200

Requirements, MDS, Audio

14411,WD Professional Archival MAF14430, Comments on Prof. Archival MAFOther topics in MAF under consideration.

Requirements Tue 1400-1800

66.6 Received National Body Comments and Liaison mattersThe NB Comments and Liaison documents for the meeting that require a response are as shown below.No. Title Response by14313 IEC CDV 61937-3 [SC 29 N 8263] None required.

14331 Liaison Statement from ETSI TC DECT to ITU-T SG 12 and ETSI TC STQ S. Quackenbush

14354 Liaison Statement from ITU-T SG 16 [SC 29 N 8324] None required.

67 Record of AhG meetingsThere were no AhG meetings prior to the 80th MPEG meeting.

68 Audio plenary, joint meeting and task group activities

218

68.1 Review of AHG reportsThere were no requests to review any of the AHG reports.

68.2 Received national body comments and liaison mattersLiaison documents were reviewed and the drafting of the responses was delegated.

68.3 Joint Meetings 68.3.1 Systems at Audio on MP4FF and Sampling Rate

David Singer, Apple, presented m14529, MP4 file format considerations for high sample-rate audio. After some discussion and further investigation done via email, it appears that all items with sampling rates greater than 2^16 -1 (65535) are written with the target value modulo 2^16. This error will be discussed during the AhG period.

68.4 Task Group discussions68.4.1 MPEG Surround

Kristofer Kjörling, Coding Technologies, presented

14453

Kristofer KjörlingJonas RödénJeroen KoppensErik SchuijersJeroen Breebaart

Proposed draft corrigendum for MPEG Surround

This contribution presented errors that make up a proposed corrigendum. These corrections are in two categories: one is proposed changes to the Enhanced Matrix Mode that result in a change in the decoded output, and the other is changes that have no impact on the decoded output.Enhanced Matrix ModeIt is proposed to change the specification so that the parameters derived from the downmix in EMM are quantized, thus permitting further processing via table look up. OtherThere are a number of editorial corrections and corrections in which the text must be changed to agree with the implementation software. The Chair noted that some of these changes do affect the bitstream syntax, but Audio Subgroup experts felt very strongly that there was no risk to fielded devices. Another correction relates to HRTF processing, which is exposed when non-symmetric HRTFs are used. The change is technically well-motivated.It is proposed that these changes be issued as a “Proposed Changes to MPEG Surround,” and possibly be issued as a DCOR at the next meeting. It was the consensus of the Audio Subgroup to incorporate all of these changes into the output document.Heiko Purnhagen, Coding Technologies, presented

14499

Johannes HilpertSascha DischHeiko PurnhagenWerner Oomen

Proposed MPEG Surround Level Enhancement

This contribution proposes the new capability of decoding of a 7.1 channel bitstream that uses a 7-2-7 structure to a 5.1 channel output, and also proposes a new level to explicitly support this case in the MPEG Surround Profile. The proposal requires some changes in the decoding specification in addition to new text describing the profile.The Audio Chair confirmed that this has been implemented in source code and there have been informal listening tests to check the implementation. It was the consensus of the Audio Subgroup to incorporate the MPEG-D changed into the “Proposed Changes to MPEG Surround,” document, and the definition of additional values of MPEG Audio profile and level into an open amendment (either 8 or 9) to MPEG-4 Audio.Heiko Purnhagen, Coding Technologies, presented

219

14504

Heiko PurnhagenAndreas SchneiderFrans de BontKarsten LinzmeierRalph Sperschneider

Proposed Updates for MPEG Surround Conformance

This contribution presents a new version of the MPEG Surround conformance document that contains the following changes and new information:

Editorial changes that account for the fact that Conformance is an amendment to MPEG Surround and not a new part.

Specification of bitsream syntax restrictions Specification of decoder conformance procedure Definition of sequence. It was noted that these sequence exist. The Chair urged the

authors to make these sequences available on some FTP site whose fpt/username/password could be publicized in an MPEG document.

68.4.2 SAOC

Hee-Suk Pang, LGE, presented

14422

Hyun-Kook LeeHee-Suk PangDong Soo KimSung-Yong YoonHenney OhYang-Won Jung

Report on the SAOC test material provided by LGE

This contribution described three proposed test items that might be used for test c). Their characteristics are summarized here:Item Number of Objects Number of Rendering Matrices

1 10 62 9 43 13 4

In every case both a mono and stereo downmix are provided.Oliver Hellmuth, FhG, presented

14441

Oliver HellmuthJuergen HerreThorsten Kastner

Proposed SAOC test items provided by Fraunhofer IIS

This contribution proposes items for SAOC tests a) and c) and for the Stream Combination test, but not for b). Specifcally:

5. items for each of tests a) and c) items for Stream Combination downmix matrices Rendering matrices

Jeroen Breebaart, Philips, presented

14464 Jeroen BreebaartWerner Oomen Proposed SAOC test items provided by Philips

This contribution proposes three items for SAOC binaural test b). They are “inside,” “telco” and “pop.” Each use the KEMAR HRTFs and can be rendered in a very flexible way using a Matlab script. This permits setting level and position parameters and these factors can also change

220

dynamically. Two “scenes” for each test item are also provided, consisting of specific downmix and rendering matrices.Heiko Purnhagen, Coding Technologies, presented

14488Jonas EngdegårdBarbara Resch

Description of SAOC test items provided by Coding Technologies

This contribution describes four sets of objects which may apply to the listening tests as shown here:

Item nr.

Test item Playback configuration and downmix specification

Rendering cases

1 Black Coffee a) R:5.1 / D:Stereo III. Complex (5 cases)2 HammerOrgan a) R:5.1 / D:Stereo I. Att./Ampl (2 cases)3 HammerOrgan a) R:5.1 / D:Stereo I. Att./Ampl (2 cases)4 VoiceOverMusic c) R: Stereo / D:Stereo I. Att./Ampl5 VoiceOverMusic c) R: Stereo / D:Stereo I. Att./Ampl6 Karaoke c) R: Stereo / D:Stereo I. Att./Ampl7 Karaoke c) R: Stereo / D:Stereo I. Att./Ampl

Jeongil Seo, ETRI, presented

14540

Seungkwon BeackJeongil SeoTaejin Leekyungok kang

Information on SAOC test items by ETRI

This contribution describes 2 candidate test items. Each has a 5.1 channel background scene object and a monoral vocal object. The items can be applied to tests a) subtest II) and III) an test b) subtest II) and III). Schuyler Quackenbush, Audio Research Labs, presented

14315 Schuyler Quackenbush

Spatial Audio Object Coding Evaluation Procedures and Criterion

This contribution is an output of the AhG on SAOC Call for Proposals. It has extensive editorial changes that improve English language usage and general organization and presentation of information. However it has yellow highlighted “to be discussed” text in several locations. These were reviewed and will be discussed later in the week.SAOC Material Selection Task On Tuesday afternoon, interested experts attended a listening task group at Apple. After a preliminary selection that day, and later in the week further listening for the binaural items via headphones, a final selection was made as shown in the table found in the following document:9099 Final Spatial Audio Object Coding Evaluation Procedures and Criterion

68.4.3 MPEG-4 ELD

Block Switching CEAll of these contributions assessed the performance of the following systems:

FhG AAC-ELD no block switching coded at 32 kb/sFT AAC-ELD no block switching coded at 32 kb/sFT AAC-ELD BS with block switching coded at 32 kb/s

The tests were done for two sets of signals, the first set containing transient material and the second having no transients (such that AAC-ELD BS never triggered a block switch).Werner Oomen, Philips, presented

14465 Erik SchuijersWerner Oomen Crosscheck FT enhanced LD AAC core experiment

221

The listening test showed that the performance of the systems under test were not different at the 95% level of significance.Markus Schnell, FhG, presented

14515

Markus SchmidtRalf GeigerMarkus Schnell

Cross-check report on Proposed FT Core Experiment for AAC-ELD

In both tests the performance of the systems under test were not different at the 95% level of significance.Henney Oh, LGE, presented

14530

Henney OhYang-Won JungHyo Jin KimChang-Heon LeeHong-Goo Kang

Cross-check report on proposed FT Core Experiment for AAC-ELD

The performance of the systems under test were not different at the 95% level of significance.Pierrick Philippe, France Telecom, presented

14519

Catherine Colomes Pierrick Philippe David Virette

Listening test results on instantaneous block switching CE for AAC ELD

This contribution presented the listening test using first the 7 items that invoke block switching. It reports that for one item, si02, FT AAC-ELD BS (with block switching) had statistically better performance at the 95% level of significance.Pierrick Philippe, France Telecom, presented

14520Pierrick Philippe David Virette

Updated description for AAC ELD instantaneous block switching CE

The contribution provided addition technical details on the operation of AAC-ELD with bock switching. It explained how the aliasing cancelling is obtained in the context of the AAC-ELD architecture, that is, both MDCT and QMF filters. The block switching introduces some slight increase in complexity, but anecdotal evidence suggests that block switching reduces the activity of TNS.Additional information was supplied, that being the listening test results pooled across all test sited doing cross-checks in this CE. For the 7 items for which block switching were active, the mean performance of FT AAC-ELD BS was higher than that of FT AAC-ELD and FhG AAC-ELD, but not at the 95% level of significance. A T-test on the difference in score between FT AAC-ELD and FhG AAC-ELD (i.e. FT AAC-ELD BS - FhG AAC-ELD and FT AAC-ELD BS - FT AAC-ELD) over the 7 items showed that this statistic was greater than zero at the 95% level of significance. As such, the T-Test revealed statistically significant improvement both on average for the 7 items under consideration and also for 4 individual items. The proposed technology has statistically similar performance for the 3 remaining items. Ralf Geiger, FhG, presented14516 Ralf Geiger Utilizing AAC-ELD for delayless mixing in

222

Markus SchnellJürgen HerreKristofer Kjörling

frequency domain

This contribution discussed the requirements for a Mixing Control Unit (MCU), particularly focussing on the requirements of low complexity and low delay. It noted that mixing in the frequency domain significantly reduces the delay through the MCU, and at the same time reduces the complexity of the mixing operation. When including the SBR filterbank, as in AAC-ELD, it is required that the SBR parameters be “merged” for the downmix signal, which is possible. DiscussionPierrick Philippe noted that having this tool in the standard does not prevent mixing in the frequency domain. In closed systems, encoders can be forced to use a given set of parameters e.g. sampling frequency, or a specific subset of tools e.g. to not use block switching.

Bernhard Grill, FhG, noted that using block switching is a “headache” for implementation, both in terms of source code for the target functionality and also in terms of encoder tuning. Therefore, he cautioned that incorporating block switching for possibly a limited quality advantage could have significant impact on coder implementation.Kristofer Kjörling noted that there is limited evidence of quality improvement, and this is balanced against concerns on complexity of implementation and use.Pierrick Philippe noted that it is very difficult to achieve statistically significant improvement for transients using the MUSHRA test methodology if they only occur for only a few frames in a waveform, but that the T test reveals such improvements.Later in the week Pierrick Philippe presented additional information, that being t-test analysis for each of the four cross-check sites. Statistical improvement with the proposed technology was revealed on the 4 test sites, no degradation were noticed for any of the items. It is Pierrick Philippe's strong opinion that this CE brings significant improvement.After considerable discussion, the Audio Chair called for a show of hands from those having strong positions on this matter. The tally was as follows:

For: 1 person from 1 company.Against: 10 persons from 4 companies

A lack of consensus for this CE was due to differing perspectives on the degree of quality improvement and the characterization of the numerous dimensions of complexity of the proposed technology (e.g. storage, computation and also implementation and coder tuning) and its applicability to identified applications.The Audio Subgroup will discuss the complexity information presented for the Block Switching CE at the 81st MPEG meeting, consulting MPEG experts from the Implementation Study Group, and agree upon metrics for balancing complexity against demonstrated quality improvement.Further Evaluation of Performance for SpeechPer Frojdh, Ericsson, presented

14501 Anisse Taleb

Report on the Evaluation of MPEG-4 Enhanced Low Delay AAC on Speech Content

This contribution showed evidence on the performance of AAC-ELD on a new test set that is more represented of speech applications. The test results showed that AAC-LD at 48 kb/s had better performance than AAC-ELD at both 38 kb/s and AAC-ELD at 32 kb/s at the 95% level of significance. Furthermore, AAC-ELD at 38 kb/s was not different from AAC-ELD at 32 kb/s at the 95% level of significance. This result agrees with the outcome of previous listening test results from France Telecom.Ralf Geiger, FhG, presented14518 Markus Schmidt

Ralf GeigerAdditional information on quality of AAC-ELD

223

Markus SchnellThis contribution showed evidence on the performance of AAC-ELD on both the MPEG-4 test set and the new test set as used in contribution m14518. The systems under test were:Codec Rate (kb/s) Delay (ms)AAC-LD 32 43AAC-ELD 32 44G.722.1-C 32 40G.722.2 (AMR-WB) 23.85 25For the speech test set, AAC-ELD was better than AAC-LD at the 95% level of significance. For the MPEG-4 test set, AAC-ELD was better than AAC-LD at the 95% level of significance.DiscussionIt was noted that, for the speech items, the FhG report test site scored AAC-ELD at 32 kb/s in the “70” range, while the Ericsson report test site scored AAC-ELD at 32 kb/s in the “80” range. This might suggest a reason for the differences Pierrick Philippe, France Telecom, volunteered to cross-check the FhG listening test result. This effort will be supported by a workplan.Erisson noted that AAC-ELD is focussed on low delay or conversational applications where speech signals are most important, but concluded that the evidence of advantage of AAC-ELD has not been confirmed by cross-check. The Chair noted that the mandate of MPEG-4 AAC-LD is coding of audio with low delay. AAC-ELD retains low delay while providing greater compression than AAC-LD for generic audio signals. Bernhard Grill, FhG, noted that in independent cross-checks using the speech items, the MUSHRA score of AAC-LD at 48 kb/s was 90 and 83, or a quality of “excellent” and the MUSHRA score of AAC-ELD at 38 kb/s was 81 and 77 or a quality at the lower range of “excellent” or upper range of “good,” and AAC-ELD at 32 kb/s was 76 and 67 or a quality of “good.” CE on low-delay SBR filterbankRalf Geiger, FhG, presented

14517

Markus SchnellJürgen HerreRalf GeigerMarkus SchmidtMarkus Multrus

Proposed Core Experiment on AAC-ELD

This contribution proposes to use a new prototype filter for the SBR filterbank that reduces the analysis/synthesis filterbank delay to 64 samples (1.3 ms) from the current SBR filterbank delay of 576 samples (12 ms). This permits the entire system one-way delay to be reduced from 42 ms to 31 ms. It presents listening test results for AAC-ELD with current SBR filterbank and with lower-delay SBR filterbank, for both high quality and low power operating modes. It was noted that there is the tendency (but not significant at the 95% level of significance) for the new filterbank to provide better performance than the original filterbank. This may be due to the asymmetry of the prototype filter which would cause little to no “pre-echo” effect.The contribution also presented filterbank frequency selectivity and filterbank computational complexity.Kristofer Kjörling, Coding Technologies, presented

14492 Fredrik Henn

Cross check of FhG Core Experiment on LD-SBR filterbank for AAC-ELD

This contribution presented the results of a cross-check listening test. The results were very similar to the FhG listening test. It was the consensus of the Audio Subgroup to accept this technology into the FPDAM text.The Audio Chair presented the following two ballot comment documents relating to ISO/IEC 14496-3:2005/PDAM 9 (AAC-ELD).

224

14286 SC 29 Secretariat

Summary of Response to Proposal of Minor Enhancement: 14496-3/Amd.9 [SC 29 N 8179]

14288 SC 29 Secretariat

Summary of Voting on ISO/IEC 14496-3:2005/PDAM 9 [SC 29 N 8180]

Concerning the first contribution, the Chair noted that a single no vote (as in this ballot) is not sufficient to delay progression of a standard.Concerning the second contribution, the Chair noted that the Finnish NB and the French NB ballot comments relate to objectives and performance of AAC-ELD, and will be further discussed. Markus Schnell, FhG, presented

14514 Markus SchnellRalf Geiger Proposed FPDAM of AAC-ELD

Heiko Purnhagen, Coding Technologies, endorsed an even simpler signalling method than what is proposed here. The Chair suggested a small break-out discuss this and report back to the group. It was the consensus of Audio Subgroup to incorporate the technology for the low-delay SBR filterbank into the FPDAM text.

68.4.4 Speech and Audio Exploration

Schuyler Quackenbush, Audio Research Labs, presented

14317 Schuyler Quackenbush

Proposed Workplan for Speech and Audio Exploration

This contribution proposed that a listening test be used to characterize the candidate test items. There was considerable disagreement as to whether this is the appropriate means to assess the test set. After some discussion, it was decided that what was most important that the test items represent significant application areas, for example streaming music, talk radio or IPTV. The Chair noted that of paramount importance is to expand the current test set. Experts will listen to all contributed items and pick new or replacement items for the test set, which will be reviewed by the Audio Subgroup.Eunmi Oh, Samsung, presented14455 Eunmi Oh Evaluation of speech and audio coding schemeThis contribution suggested guidelines for listening tests associated with assessing signals that are mixed signals, e.g. both speech and audio. Specifically, that participants listen to stimuli three times: once to assess e.g. speech coding artefacts, once to assess music coding artefacts and once to assess how the two categories of impairment could be combined to form an overall judgement. It was also noted that items of duration of not more than 15 seconds would be best considering that listeners should listen to them numerous times.Additionally, the contribution described three new mixed-signal items that Samsung has contributed to the set of candidate items.

68.4.5 Symbolic Symbol Representation

Pierfrancesco Bellini, UNFI, presented

14364

Pierfrancesco BelliniPaolo NesiMaurizio CampanaiGiorgio Zoia

Editors study on ISO/IEC 14496-23/FCD

The contribution is candidate text for the FDIS test to be produced at this meeting. All changes are in response to ballot comments from the UK, Italian and Korean NBs. The Chair noted that the SMR editors should consider some demonstration or publicity vehicle that might play the role of a verification test and hence server to demonstrate to MPEG and the larger community of

225

potential customers the range of functionalities supported by SMR. As a minimum, this could include a technology demonstration at the closing MPEG plenary at the Lausanne meeting.The SMR task group members:

integrated the comments received for the korean ornaments definition integrated the changes proposed in contribution m14364 prepared the DoC

68.4.6 MPEG-1, MPEG-2 and MPEG-4 audio, conformance, reference software

Werner Oomen, Philips, presented

14536

Frans de BontWerner Oomen

Cor to 14496-3:2005 subpart 10, DST (lossless oversampled audio)

The contribution proposes corrections to the DST specification. This an error that occurs in two places and which permits the specification to support a greater number of channels. This will issue as a DCOR from this meeting.Kelvin Lee, I2R, presented

14414Kelvin LeeTe LiHaibin Huang

Proposed Corrigenda to 14496-3:2005/AMD 3 (SLS)

This contribution corrects an error that appears a number of places in the text relating to the sign of the residual. It also corrects values in a number of tables.Mauri Vannen, Nokia, presented

14522 Juha Ojanperä miikka.vilermo@nokia.com Miikka Vilermo

On AAC LTP conformance

There was considerable discussion on the issues that made conformance testing of LTP a difficult problem. The Chair encouraged Nokia experts to maintain their momentum in this effort and to propose at the next meeting:

A conformance procedure Conformance bitstreams Informative text on encoder operation strategies that would produce bitstreams that, when

decoded, always meet conformance criterion.

Noboru Harada, NTT, presented

14410

Noboru HaradaTakehiroMoriyaYutaka Kamamoto

Proposed revision for ISO/IEC14496-3, AMD8: MP4FF box for original audio file information

This text will have some additional edits

Ralph Sperschneider, FhG, discussed14355 Ralph Sperschneider WD on MPEG-4 Audio Fourth EditionThe Chair urged all experts to review this text. It will be output as WD from this meeting as we wish to incorporate AMD 9 (BSAC and SBR) into this edition, and the final ballot for AMD 9 has not yet closed.Tilman Liebchen, LGE, presented

14428 Tilman Liebchen

Proposed Text of ISO/IEC 14496-4:2004/FDAM 19, Audio Lossless Coding (ALS) Conformance

14429 Tilman Liebchen Updated Status of ALS Conformance

226

These two contributions are revised text for Conformance FDAM 19 and also an update on the status of ALS conformance. Currently all bitstreams are defined, available and cross-checked.Kelvin Lee, I2R, presented14407 Kelvin Lee Status of SLS reference software updateThis contribution reports that the “stand-alone” SLS reference software now supports MP4FF, and that AAC LC BSAC can be used as a core coder in mono, stereo and multichannel and that SLS operates in non-core mode in mono, stereo and multichannel.The Chair suggested that FhG and I2R work together during the next AhG to define an API such that the stand-alone code could be linked with the MP4VM so as to be part of the unified framework, and that they report back at the next MPEG meeting as to whether this integration method is feasible.

69 Meeting deliverables69.1 Recommendations for final plenaryThe Audio recommendations were presented and approved.

69.2 Establishment of Ad-hoc GroupsThe following ad-hoc groups were established by the Audio subgroup:

No. Title Mtg9097 AHG on Audio Standards Maintenance No9098 AHG on SAOC CfP, Speech and Audio and AAC-ELD Yes

69.3 Approval of output documentsAll output documents, shown in Annex D, were presented in Audio plenary and were approved.

69.4 Responses to Liaison and NB commentsThe responses to Liaison and NB comments were prepared and approved.

69.5 Press statementThe Audio part of the press statement was prepared and approved.

70 Future activities70.1 Schedule of future meetingsAd Hoc group meetings are indicated in Section 69.2. Unless otherwise indicated, Ad Hoc group meetings will be held at the location of the next MPEG meeting on the weekend preceding that meeting.

70.2 Agenda for next meetingThe agenda for the next MPEG meeting is shown in Annex E.

70.3 All other businessThere was none.

70.4 Closing of the meeting The 80th Audio Subgroup meeting was adjourned Friday at 14:00.

227

Annex A Participants

First Name Last NameCountry Affiliation

Pierfrancesco Bellini Italy DSI-UNIFIJeroen Breebaart NL PhilipsKok Seng Chong SG PanasonicMatt Fellers USA DolbyRalf Geiger DE Fraunhofer IISMatthias Gruhne DE FhG IIS AEMTNoboru Harada JP NTTOliver Hellmuth DE Fraunhofer IISJürgen Herre DE Fraunhofer IISHaibin Huang SG I2RYang-Won Jung KR LG ElectronicsDong Soo Kim KR LG Electronics

Kristofer Kjörling SCoding Technologies

Kelvin Lee SG I2RTe Li SG I2RTilman Liebchen DE LG ElectronicsTakehiro Moriya JP NTTMarkus Multrus DE Fraunhofer IISSua Hong Neo SG PanasonicToshiyuki Nomura JP NECTakeshi Norimatsu JP PanasonicEunmi Oh KR SamsungHenney Oh KR LG Electronics

Werner Oomen NLPhilips Applied Technologies

Hee-Suk Pang KR LG Electronics

Pierrick Philippe FRFrance Telecom R&D

Heiko Purnhagen SECoding Technologies

Schuyler Quackenbush USA ARLSusanto Rahardja SG I2R

Jonas Rödén SECoding Technologies

Juergen Schmidt DE Thomson

Andreas Schneider DECoding Technologies

Markus Schnell DE Fraunhofer IISJeongil Seo KR ETRIOsamu Shimada JP NECRalph Sperschneider DE Fraunhofer IISMauri Vaananen FIN Nokia Res. Center

Jyri Huopaniemi FINNokia Research Center

Do-Hyung Kim KR Samsung AIT

Annex B Audio Contributions and Schedule

Monday

0900-1200 MPEG Plenary

1200-1400 Lunch

1400-1800 Audio Plenary

Welcome

Approval of previous meeting report

14316 Schuyler Quackenbush 79th MPEG Audio Report

AhG Reports

14281 R. Sperschneider AHG on Audio Standards Maintenance

14282 S. Quackenbush AHG on SAOC CfP and AAC-ELD

Liaison

14313 IEC TC 100 via SC 29 Secretariat IEC CDV 61937-3 [SC 29 N 8263]

14331 ETSI TC DECT via SC 29 Secretariat

Liaison Statement from ETSI TC DECT to ITU-T SG 12 and ETSI TC STQ

14354 ITU-T SG 16 via SC 29 Secretariat

Liaison Statement from ITU-T SG 16 [SC 29 N 8324]

Ballot comments

14286 SC 29 Secretariat Summary of Response to Proposal of Minor Enhancement: 14496-3/Amd.9 [SC 29 N 8179]

14287 SC 29 Secretariat Summary of Voting on ISO/IEC TR 11172-5:1998/DCOR 1 [SC 29 N 8178]

14288 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2005/PDAM 9 [SC 29 N 8180]

14292 ITTF via SC 29 Secretariat Table of Replies on ISO/IEC 14496-3:2005/FDAM 1 [SC 29 N 8207]

14319 SC 29 Secretariat Summary of Voting on ISO/IEC 13818-7:2006/FPDAM 1 [SC 29 N 8268]

14320 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 18 [SC 29 N 8269]

14321 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 19 [SC 29 N 8270]

14327 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-4:2004/FPDAM 14 [SC 29 N 8276]

14328 SC 29 Secretariat Summary of Voting on ISO/IEC FCD 14496-23 [SC 29 N 8277]

14344 SC 29 Secretariat Summary of Voting on ISO/IEC 23003-1/PDAM

1 [SC 29 N 8307]

14345 SC 29 Secretariat Summary of Voting on ISO/IEC 23003-1/PDAM 2 [SC 29 N 8308]

14384 SC 29 Secretariat Summary of Voting on ISO/IEC 14496-3:2005/PDAM 8

MPEG Surround

14453

Kristofer KjörlingJonas RödénJeroen KoppensErik SchuijersJeroen Breebaart

Proposed draft corrigendum for MPEG Surround

14499

Johannes HilpertSascha DischHeiko PurnhagenWerner Oomen

Proposed MPEG Surround Level Enhancement

14504

Heiko PurnhagenAndreas SchneiderFrans de BontKarsten LinzmeierRalph Sperschneider

Proposed Updates for MPEG Surround Conformance

SAOC

14422

Hyun-Kook LeeHee-Suk PangDong Soo KimSung-Yong YoonHenney OhYang-Won Jung

Report on the SAOC test material provided by LGE

14441Oliver HellmuthJuergen HerreThorsten Kastner

Proposed SAOC test items provided by Fraunhofer IIS

14464 Jeroen BreebaartWerner Oomen Proposed SAOC test items provided by Philips

14488 Jonas EngdegårdBarbara Resch

Description of SAOC test items provided by Coding Technologies

14540

Seungkwon BeackJeongil SeoTaejin Leekyungok kang

Information on SAOC test items by ETRI

14315 Schuyler Quackenbush Spatial Audio Object Coding Evaluation Procedures and Criterion

Tuesday

0900-1300 AAC-ELD

230

14465 Erik SchuijersWerner Oomen

Crosscheck FT enhanced LD AAC core experiment

14515Markus SchmidtRalf GeigerMarkus Schnell

Cross-check report on Proposed FT Core Experiment for AAC-ELD

14530

Henney OhYang-Won JungHyo Jin KimChang-Heon LeeHong-Goo Kang

Cross-check report on proposed FT Core Experiment for AAC-ELD

14519Catherine Colomes Pierrick Philippe David Virette

Listening test results on instantaneous block switching CE for AAC ELD

14520 Pierrick Philippe David Virette

Updated description for AAC ELD instantaneous block switching CE

14516

Ralf GeigerMarkus SchnellJürgen HerreKristofer Kjörling

Utilizing AAC-ELD for delayless mixing in frequency domain

14501 Anisse Taleb Report on the Evaluation of MPEG-4 Enhanced Low Delay AAC on Speech Content

14518Markus SchmidtRalf GeigerMarkus Schnell

Additional information on quality of AAC-ELD

1300-1400 Lunch

1400-1600 SAOC Material Selection (at Apple)

1400-1800 Joint meeting with Requirements, MDS, Audio at Requirements

14411,WD Professional Archival MAF14430, Comments on Prof. Archival MAFOther topics in MAF under consideration.

1800-1900 Liaison Meeting

Response to 14331, ETSI TC DECT

1900- Chairs Meeting

Wednesday

0900-1100 MPEG Plenary

1130-1200 Joint with Systems at Audio

14529 David Singer MP4 file format considerations for high sample-rate audio

Discuss Ballot Comments on MP4FF box

1200-1300 Speech and Audio Exploration

231

14317 Schuyler Quackenbush Proposed Workplan for Speech and Audio Exploration

14455 Eunmi Oh Evaluation of speech and audio coding scheme

1300-1400 Lunch

1400-1500 AAC-ELD

14492 Fredrik Henn Cross check of FhG Core Experiment on LD-SBR filterbank for AAC-ELD

14517

Markus SchnellJürgen HerreRalf GeigerMarkus SchmidtMarkus Multrus

Proposed Core Experiment on AAC-ELD

14514 Markus SchnellRalf Geiger Proposed FPDAM of AAC-ELD

1500-1530 SMR

14364

Pierfrancesco BelliniPaolo NesiMaurizio CampanaiGiorgio Zoia

Editors study on ISO/IEC 14496-23/FCD

1530-1730 MPEG-4

14536 Frans de BontWerner Oomen

Cor to 14496-3:2005 subpart 10, DST (lossless oversampled audio)

14414Kelvin LeeTe LiHaibin Huang

Proposed Corrigenda to 14496-3:2005/AMD 3 (SLS)

14522Juha Ojanperä miikka.vilermo@nokia.com Miikka Vilermo

On AAC LTP conformance

1730- Social

Thursday

0900-1000

14410Noboru HaradaTakehiroMoriyaYutaka Kamamoto

Proposed revision for ISO/IEC14496-3, AMD8: MP4FF box for original audio file information

14355 Ralph Sperschneider WD on MPEG-4 Audio Fourth Edition

14428 Tilman Liebchen Proposed Text of ISO/IEC 14496-4:2004/FDAM 19, Audio Lossless Coding (ALS) Conformance

14429 Tilman Liebchen Updated Status of ALS Conformance

14407 Kelvin Lee Status of SLS reference software update

1030-1300 Break-out Task Group Activity

232

SAOC EvaluationSAOC Binaural material selectionSpeech and Audio material selection

1300-1400 Lunch

1400- SAOC Evaluation Document

1730-1800 Approve Liaison Responses

1800- Chairs Meeting

Friday Audio plenary

0900-1300 Recommendations for final plenary

Establishment of new Ad-hoc groups

AhG Mandates

Get document numbers

1000 Approve Responses to NB comments

1030 Approval of output documents

Review of Audio presentation to MPEG plenary

Agenda for next meeting

A.O.B.

Closing of the Audio meeting

1300-1400 Lunch (optional!)

1400- MPEG Plenary

233

Annex C Task Groups

1. MPEG-D MPS

2. MPEG-D SAOC

3. MPEG-4 AAC-ELD

4. Speech and Audio

5. MPEG-1 reference software

6. MPEG-2 audio

7. MPEG-4 audio, conformance, reference software

Annex D Output DocumentsNo. Title TBP Available

11172-5 Software simulation9064 DoC on ISO/IEC 11172-5:199x/DCOR 1 No 07-04-279065 ISO/IEC 11172-5:199x/Cor. 1 No 07-04-27

13818-7 Adavnced Audio Coding9066 DoC ISO/IEC 13818-7:2006/FPDAM 1 No 07-04-27

9067 ISO/IEC 13818-7:2006/FDAM 1, Transport of MPEG Surround data in AAC

No 07-04-27

14496-3 Audio9068 ISO/IEC 14496-3:2005/DCOR 5 (DST and MP3on4) No 07-04-279069 ISO/IEC 14496-3:2005/DCOR 6 (SLS) No 07-06-089070 DoC on ISO/IEC 14496-3/PDAM 8 No 07-04-27

9071 ISO/IEC 14496-3/FPDAM 8, MP4FF Box for Original Audio File Information

No 07-04-27

9072 DoC on ISO/IEC 14496-3:2005/PDAM 9 Request for Amendment. No 07-04-279073 DoC on ISO/IEC 14496-3:2005/PDAM 9 No 07-04-279074 ISO/IEC 14496-3:2005/FPDAM 9, AAC-ELD No 07-06-089075 WD on MPEG-4 Audio Fourth Edition No 07-06-08

14496-4 Conformance testing9076 DoC on ISO/IEC 14496-4:2004/FPDAM 14 No 07-04-279077 ISO/IEC 14496-4:2004/FDAM 14, BSAC Extensions Conformance No 07-04-279078 DoC ISO/IEC 14496-4:2004/FPDAM 18 No 07-04-27

9079 ISO/IEC 14496-4:2004/FDAM 18, MPEG-1 and -2 on MPEG-4 Conformance

No 07-04-27

9080 DoC ISO/IEC 14496-4:2004/FPDAM 19 No 07-04-279081 ISO/IEC 14496-4:2004/FDAM 19, ALS Conformance No 07-04-279082 Study on ISO/IEC 14496-4:2004/FPDAM 20, SLS Conformance No 07-04-279083 Status of MPEG-4 Audio Conformance No 07-04-279084 Status of MPEG-4 SLS Conformance No 07-04-27

14496-5 Reference Software 9085 ISO/IEC 14496-5:2001/AMD 10:2007/DCOR 1, BSAC and SLS No 07-04-279086 Request for Amendment, MPEG-1/2 on MPEG-4 Ref. Software No 07-04-27

9087 ISO/IEC 14496-5:2001/AMD 20, MPEG-1/2 on MPEG-4 Ref. Software

No 07-04-27

14496-23 Symbolic Music Representation9088 DoC ISO/IEC FCD 14496-23 No 07-05-119089 ISO/IEC FDIS 14496-23:200x, Symbolic Music Representation No 07-05-11

23003-1 MPEG Surround

9099 Final Spatial Audio Object Coding Evaluation Procedures and Criterion

No 07-04-27

9090 DoC ISO/IEC 23003-1:2007/PDAM 1 No 07-04-279091 ISO/IEC 23003-1:2007/FPDAM 1, MPEG Surround Conformance No 07-06-089092 DoC ISO/IEC 23003-1:2007/PDAM 2 No 07-04-27

9093 ISO/IEC 23003-1:2007/FPDAM 2, MPEG Surround Reference Software

No 07-06-08

9094 Defect Report of ISO/IEC 23003-1:2007 No 07-04-27Audio and speech coding

9095 Framework for Exploration of Speech and Audio Coding No 07-04-27

235

9096 Workplan for Exploration of Speech and Audio Coding No 07-04-27

236

Annex E Agenda for the 81st MPEG Audio Meeting

Agenda Item1. Opening of the meeting2. Administrative matters

2.1. Approval of agenda and allocation of contributions2.2. Communications from the Chair2.3. Joint meetings2.4. Review of task groups and mandates2.5. Approval of previous meeting report2.6. Review of AhG reports 2.7. Received national body comments and liaison matters

3. Plenary issues4. Task group activities

4.1. MPEG Maintenance, including MPEG-1, MPEG-2, MPEG-4, SMR and MPEG Surround issues

4.2. AAC-ELD4.3. Spatial Audio Object Coding Call for Proposals Evaluation4.4. Speech and Audio Exploration

5. Discussion of unallocated contributions6. Meeting deliverables

6.1. Recommendations for final plenary6.2. Establishment of new Ad-hoc groups6.3. Approval of output documents6.4. Responses to NB comments6.5. Responses to Liaison statements6.6. Press statement

7. Future activities8. Agenda for next meeting9. A.O.B10. Closing of the meeting

237

Annex K – 3DG report

Source: MPEG 3D Graphics CompressionTitle: 3D Graphics Marrakech meeting reportAuthors: Marius Preda (INT)Status: Draft (To be added to Nxxxx)

3DG meeting reportSan Jose, April 23-28, 2007

1 Opening of the Meeting

70.5 Approval of the agenda

70.6 Goals for the weekThe goals of this week are:

Review FAMC results and edit the WD Review on-going AFX experiments Promote the 3DGC profiles Issue FDAM of GFX reference software Issue FDAM of GFX conformance Issue FPDAM of Geometry and Shadow reference software Issue FPDAM of Geometry and Shadow conformance Review Liaisons to MPEG 3DG Review and promote 3DG related demonstrations Investigate future developments of MPEG 3D Graphics3

The output documents related to 3D Graphics Compression are:

No. Title TBP Available Editor14496-4 Conformance testing

9132 Text of ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

No 07/05/12 Mark Callow

9146 DoC on ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

No 07/05/12 Marius Preda

9133 Text of ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

No 07/04/27 Jeong-Hwan Ahn

9147 DoC on ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

No 07/05/04 Marius Preda

No. Title TBP Available Editor14496-5 Reference Software

9134Text of ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)PB: Mark has to clean up the code

N 07/05/12 Mark Callow

238

9148 DoC on ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)

N 07/05/12 Marius Preda

9135

Text of ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)PB: Patrick has to send me the software from the CVS

N 07/05/04 Patrick Gioia

9149 Doc on ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)

N 07/05/04 Marius Preda

No. Title TBP Available Editor14496-16 Animation Framework eXtension (AFX)

9136 WD 2.0 of ISO/IEC 14496-16:2006/AMD2 (Frame-based Animated Mesh Compression)

N 07/04/27 Marius Preda, Titus Zaharia

9137 WD 1.0 of ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

N 07/04/27 Patrick Gioia

9150 Request for ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

N 07/04/27 Marius Preda

9138 3D Graphics Core Experiments Description N 07/04/27 Khaled Mammou9139 3D Graphics Compression FAQ 19.0 Y 07/05/12 Pierre Davy

No. Title TBP Available Editor14496-21 MPEG-J GFX

9140 Text of ISO/IEC 14496-21:2006/COR1 N 07/04/27 Mark Callow

No. Title TBP Available Editor14496-25 3D Graphics Compression Model

9141 Request for Subdivision of ISO/IEC 14496: Part 25 - 3D Graphics Compression Model

N 07/04/27 Marius Preda

9142 WD 1.0 for ISO/IEC 14496-25 Y 07/04/27 Marius Preda

70.7 Standards from 3DG

Std Pt Edit. Project Description CfP WD CDPDAMDCOR

FCDFPDAM

FDISFDAMCOR

4 4 2004 Amd.16 MPEG-J GFX conformance

06/04 06/10 07/04

4 4 2004 Amd.21 Geometry and Shadow conformance

06/07 06/10 07/04 07/10

4 5 2001 Amd.11 MPEG-J GFX reference software

06/01 06/04 06/07 07/04

4 5 2001 Amd.13 Geometry and Shadow reference software

06/07 06/10 07/04 07/10

4 16 2006 Amd.1 Geometry and Shadow

05/04 06/04 06/07 07/01

4 16 2006 Amd.2 Frame-based 07/01 07/07 07/10 08/04

239

Animated Mesh Compression

4 21 2006 Cor.1 07/01 07/044 25 200x 3D Graphics

Compression Model07/10 08/01 08/07

70.8 Room allocation3DG : Santa Clara

70.9 Allocation of contributions

N° Title Schedule ActivityD1 Monday D1

MPEG Plenary D1 09:00~11:30 MPEG General

3DG Plenary D1 11:30~13:00 3DG General

Roll call, Agenda, Goals, FAQ, etc. Marius Preda

14269Report of AHG on 3DGC documents, experiments and software maintenance

Francisco MoránJeong-Hwan AhnMark Callow

Lunch Break D1 13:00~14:00

Web Site D1 14:00~14:15

Clarify the status on www.mpeg-3dgc.org maintenance all

Conformance D1 14:15~14:30 Conformance

14396 Conformance bitstream for Geometry & Shadow

Jeong-Hwan Ahn

Reference Software D1 14:30~14:45

Clarify the status on node templates and Stream Code all

GFX D1 14:45~15:00

Report on Reference Software and Corrigendum status (latest developments, demo)

Mark Callow

New issues D1 15:00~15:30

14545 A scene graph node designed to define haptic properties

Pierre DavyNadia Magnenat-Thalmann

240

N° Title Schedule Activity

Coffee Break 15:30~16:00

New issues D1 16:00~16:30

Proposal for Future developments in MPEG 3D Graphics

Marius Preda

Requirements D1 17:00~17:50

14467 Proposal for 3D Compression Profile

Patrick GioiaOlivier Aubault

Preliminary Discussion

D2 Tuesday D2

Core Experiments D2 09:00~9:45 CE2

14466 Report on CE2: Space Partitioning

Patrick GioiaAnne Le BrisRomain Cavagna

Core Experiments D2 09:45~10:30 CE1

14363 Scalable Compression of Dynamic 3D Meshes

Nikolce Stefanoski Jörn Ostermann

CE1

Coffee Break 10:30~11:00

14493Frame-based Animated Mesh Compression : integration of the CABAC arithmetic encoder

Khaled MamouKarsten MüllerDetlev MarpeTitus ZahariaMarius PredaFrancoise Prêteux

CE1

14498 FAMC with streaming support

Khaled MamouTitus ZahariaMarius PredaFrançoise Prêteux

CE1

14491 FAMC bitstream description

Khaled MamouMarius PredaTitus ZahariaFrancoise Prêteux

CE1

Lunch Break D2 12:00~14:00

Core Experiments D2 14:00~14:30 CE1

241

N° Title Schedule Activity

14491 FAMC bitstream description

Khaled MamouMarius PredaTitus ZahariaFrancoise Prêteux

CE1

Miscellaneous D2 15:20~15:30

14408 3dod.org goes multimedia: MyMultimediaWorld.com

Marius Preda Benoit Le Bonhomme Son Tran Françoise Preteux

New issues D1 15:00~16:00

Proposal for Future developments in MPEG 3D Graphics

Marius Preda

Liaison D2 16:00~17:00

Liaison Statements LiaisonD3 Wednesday D3

MPEG Plenary D3 09:00~12:00

MPEG General

Joint meeting with Requirements D2 12:00~12:30 Profile

14467 Proposal for 3D Compression Profile

Patrick GioiaOlivier Aubault

Lunch Break D3 12:30~14:00

3DG Plenary D3 14:00~17:00 3DG General

WD 2.0 Editing

14396 Conformance bitstream for Geometry & Shadow (step 2)

Jeong-Hwan Ahn

Clarify the status on node templates and Stream Code all

D4 Thursday D4

Lunch Break D4 12:00~14:00

3DG documens D4 14:00~18:00 3DG General

GFX Output documents reviewCore Experiment discussionCE 1 ReviewCE 2 ReviewAMD2 3D Multiresolution ProfileOutput documents review

242

N° Title Schedule ActivityD5 Friday D5

3DG documens D4 09:00~12:00 3DG General

Output documents reviewShort Report on the Crosschecking statusShort report on the FAQAMD 3 Profile

AhGs and resolutions

Lunch Break D5 12:00~14:00

MPEG Plenary D5 14:00~ MPEG General

70.10 Attendance list

Name Country Company e-mailJeong-Hwan Ahn Korea Samsung AIT jeonghwan.ahn @ samsung . com Marius Preda France INT marius.preda @ int-evry . frFrançoise Prêteux France INT Francoise.Preteux @ int-evry . frKhaled Mamou France INT Khaled.Mamou @ int-evry . frPatrick Gioia France France Telecom

R&Dpatrick.gioia @ orange-ftgroup . com

Euee S. Jang Korea Hanyang Univ. esjang @ hanyang . ac . krSunyoung Lee Korea Hanyang Univ. sunnykr @ ihanyang . ac . krSinwook Lee Korea Hanyang Univ. nembi79 @ gmail . comJae Bum Jun Korea Hanyang Univ. powerory @ hanyang . ac . krHyungyu Kim Korea Hanyang Univ. cprov @ cpsite . netDan Cernea Belgium VUB cdcostin @ etro . vub . ac . beMark Callow Japan HI Corporation callow_mark @ hicorp . co . cpKarsten Muller Germany FHG-HHI kmuller @ hhi . dePierre Davy Swiss Miralab davy @ miralab.unige.chNing Lu US Intel

CorporationJörn Ostermann Germany Institut für

Informationsverarbeitung

Ostermann @ tnt.uni-hannover.de

Anne Le Bris France France Telecom anne.lebris @ simecom.fr

243

71 General issues

71.1 General Discussion

71.1.1 ExperimentsLast meetings resolutionFor each new specification development activity, 5 National Bodies should commit resources to that activity. Contributions should be made at each meeting from those NBs until that activity is finalized.3DGC will no longer have Exploration Experiments.3DGC will only have Core Experiments for any official experiments.The condition for the CE is to have at least 2 active participants (companies or universities having support from companies on that experiment) dedicating resources to do the work and making contributions at each meeting.If a participant does not make any contribution at a meeting, then that participant will not be considered as active.The activity in the CE does not necessarily imply adoption into the standard.

xxx Clarify the status on www.mpeg-3dgc.org maintenance all

Samsung can maintain the web site up to end of 2007 only. Potential solutions: FT and UPM. Patrick Gioia will be the maintainer of the new web site (once transferred). FT will investigate on transferring the web site and on finding open source solutions for data protection.

14396 Conformance bitstream for Geometry & Shadow

Jeong-Hwan Ahn

A table with the responsible person for cross-checking was created. Some of the files are not yet provided (Multiresolution FootPrint). It is possible that same files show functionalities in the two table. Jerome (FT) will check and if not he’ll provide new files. The issue will be re-discussed during the week.

xxx Clarify the status on node templates and Stream Code all

Stream Code problem was solved by correcting the Geometry and Shadow spec and updating the RefSoft accordingly. Jerome (FT) will provide a new version of templates8.txt document.

xxxReport on Reference Software and Corrigendum status (latest developments, demo)

Mark Callow

244

Reference Software is in good form but still needs some clean up (this will be done in the editing period). Demonstration of providing the Java MIDP environment with GFX API implementation was shown. Demonstration material is provided as RefSoft and Conformance

14545 A scene graph node designed to define haptic properties

Pierre DavyNadia Magnenat-Thalmann

Proposal of a new node in the scene graph.Examples of using haptics devices: game, touching virtual objects, medical training, interface for content production. The problem that has to be solved is how to compute the force to be directed to the device based on haptics properties of the virtual 3D graphics object. Software solutions exist: direct communication with the device, haptic geometry, extract the geometry from 3d models. The contribution brings a proposition on the parameters to be attached to the graphics object.In order to take a decision to start the CE evidences have to be provided for a support from industrial partners. Also the requirement of treating this kind data in MPEG has to be established.

Resolution: Proponents are asked to provide more evidences that such tool is currently required by the industry.

14408 3dod.org goes multimedia: MyMultimediaWorld.com

Marius Preda Benoit Le Bonhomme Son Tran Françoise Preteux

This contribution presents the latest developments of the web site 3dod.org (now called MyMultimediaWorld.com) showcasing AFX tools.

xxx Proposal for Future developments in MPEG 3D Graphics

Marius Preda

A new architecture on considering 3DGC tool was presented. It is based on three layers structure: XML-based representation for scene graph, Generic Binarization of XML content Specific Compression tools for 3D Graphics Primitives

The group acknowledged the advantages of such approach in promoting the AFX tools to the industry.

14467 Proposal for 3D Compression Profile Patrick GioiaOlivier Aubault

Preliminary Discussion

245

The contribution presented an improved version of the Multiresolution Profile (profile under consideration from the last meeting).Issues : having all the tools in the profile and control it by using the levels or selecting only the tools are really needed.The levels should be specified for each tool. This issue will be revised during the week.

14466 Report on CE2: Space Partitioning

Patrick GioiaAnne Le BrisRomain Cavagna

This contribution presents the results of the exploratory phase. The goal was to specify a sound framework for space partitioning that may work for all toolsAn initial representation for PVS and Cell and Portals is presented. A more compact form should be provided. Next stet of the CE is the competitiveness phase: design an efficient data representation. Participants are ENST and FT.

14363 Scalable Compression of Dynamic 3D Meshes

Nikolce Stefanoski Jörn Ostermann

CE1

The contribution presents a method for scalable representation of the geometry and animation for all layers. The compression results with respect to FAMC are presented.

14493Frame-based Animated Mesh Compression : integration of the CABAC arithmetic encoder

Khaled MamouKarsten MüllerDetlev MarpeTitus ZahariaMarius PredaFrancoise Prêteux

CE1

The contribution presents the adaptation of the CABAC for FAMC. Introduction of the CABAC as it is used in video. The new results show an improvement of 15%.

14498 FAMC with streaming support

Khaled MamouTitus ZahariaMarius PredaFrançoise Prêteux

CE1

The contribution presents the partition of the FAMC stream for enabling animation streaming. The skinning model may be computed for each segment. For some examples, doing so improves the bit-rate.

246

14491 FAMC bitstream description

Khaled MamouMarius PredaTitus ZahariaFrancoise Prêteux

CE1

This contribution presents the bitstream syntax of the FAMC. It includes the new development for streaming and CABAC integration.

14467 Proposal for 3D Compression Profile Patrick GioiaOlivier Aubault

Final Discussion

This contribution is presented in the joint meeting 3DGC-Requirements. Presentation of the compression tools to be supported in the profile. Accepted as a new AMD of ISO/IEC 14496-16.

14396 Conformance bitstream for Geometry & Shadow (step 2)

Jeong-Hwan Ahn

All the bitstreams are available in www.mpeg-3dgc.org databank.

xxx Clarify the status on node templates and Stream Code all

NodeTemplatev8.txt is updated and available on CVS.

WD 2.0 Editing 14:00-18:00

Technical review was performed. Pictures have to be updated.

GFX Output documents review

The ISO/IEC 14496-21:2006COR was updated. Add a new method for binding textures. Change the name of a class.

The ISO/IEC 14496-5: FDAM 11 was updated.

The ISO/IEC 14496-4: FDAM 16 was updated.

72 AFX (14496-16) activities

247

72.1 Core Experiments

72.1.1 CE1. Mesh Animation CompressionLast meeting resolutionContinue CE1 with the next steps (representation, compression of other attributes and considering static and animated data together).Issue a working draft with the currently proposed technology.Issue a request for new amendment document with the title “Frame-based Animated Mesh Compression”

72.1.1.1 M14493 –Frame-based Animated Mesh Compression : integration of the CABAC arithmetic encoder

This proposal describes an approach of integrating CABAC into the FAMC technology. As was shown by experimental results, this proposed enhancement of FAMC results in average bit-rate savings of around 16% when compared to the current WD. At the same time, by replacing the N-ary adaptive arithmetic coder in the current WD by the fast multiplication-free M coder, as being an integral part of CABAC, computational complexity was reduced.

72.1.1.2 M14498 – FAMC with streaming support

This proposal describes a data packetization mechanism that enables the FAMC technique with the streaming functionality. The proposed approach makes it possible to associate multiple skinning models with a single animation sequence and therefore to optimize the motion model to each data segment. The experimental results, carried out on the 3DGC test data set, established that the streaming can be efficiently performed, with marginal loss in term of compression efficiency.

72.1.1.3 M14491 – FAMC bitstream description

This proposal describes new bistream description for FAMC including the changes from the previous two contributions.

72.1.1.4 M14363 – Scalable Compression of Dynamic 3D Meshes (SCD3DM)

This proposal describes a method for predictive compression of time-consistent 3D mesh sequences supporting and exploiting scalability. The applied method decomposes each frame of a mesh sequence in layers, which provides a time-consistent multi-resolution representation. Following the predictive coding paradigm, local temporal and spatial dependencies between layers and frames are exploited for compression. Prediction is performed vertex-wise from coarse to fine layers exploiting the motion of already encoded neighboring vertices for prediction of the current vertex location. It is shown that a predictive exploitation of the proposed layered configuration of vertices can improve the compression performance in domains relevant for applications.

Discussion on CE1

248

The compression results presented in M14363 (SCD3DM) are generally comparable with the ones in FAMC. However FAMC performs better for low-bitrates. SCD3DM introduces the animation on different geometry resolution.

Resolution for CE1In the next phase of the CE it will be investigated how the skinning model will be combined with the scalable approach. A switch may be used to choose between DCT, Wavelet compression of the errors and the scalable approach.Issue a new version of the working draft with the currently proposed technology (including streaming and CABAC).

72.1.2 CE2. Space PartitioningLast meeting resolutionPerform exploratory stage with the proposed work plan. (details can be found in the CE description)

72.1.2.1 M14466 – Report on CE2: Space Partitioning

This contribution presents the results of the exploratory phase. The initial goal of specifying a sound framework for space partitioning that may work for all tools (PVS, BSP, Cell and Portal) was achieved. Reference implementation and data test was provided. Discussion on CE1Next step of the CE is the competitiveness phase with the goal of design an efficient data representation. Participants are ENST and FT.

Resolution for CE2Perform competitiveness stage with the proposed work plan. (details can be found in the CE description)

72.2 Profiles

72.2.1 M14467 – Proposal for 3D Navigation ProfileAfter two stages review (internal and joint with Requirements) it was acknowledged that the proposal is mature to start the publishing stage.

ResolutionRequest a new AMD and prepare the first draft.

72.3 Promotions

72.3.1 M14408 – 3dod.org goes multimedia: MyMultimediaWorld.com

249

The goal of this contribution is to present the evolution of the 3dod.org for becoming a multimedia repository showcasing MPEG-4 technology for representing and delivering the content. It supports on-line visualization of 3D graphics, video, image and sound content, categories and user management, content upload and conversion, content adaptation.

73 GFX (14496-21) activities

73.1 Reference Software & ConformanceLast meeting resolutionThe proposed restructuring of the reference software is approved.The video files shall be replaced by the next meeting.In order to synchronize the reference software with conformance schedule, this document will be promoted to FDAM at the next (80th) meeting.Study document of the DoC and the Text will be provided at this meeting.

73.1.1 M14091 – Summary of Voting on ISO/IEC 14496-5:2001/FPDAM 11 (from last meeting)The JNB has voted disapprove with two comments.The first comment is to restructure the reference software to support J2ME. Although it is classified as technical comment, it is only an implementation issue and thus is considered as editorial fix by the group. However, this will affect the conformance work which is scheduled to be promoted to FDAM stage at this meeting. Therefore, the group approves the comment but recommends synchronizing the work with the conformance schedule.

The second comment is to remove the video files that are used to demonstrate the reference software because of license issue. However, since another movie file should be provided, the group approves to replace (not remove) the video files

ResolutionThe proposed restructuring of the reference software is approved.The document was promoted to FDAM.DoC and the Text were provided at this meeting.

74 3D Graphics Compression Model (14496-25) activities

A new architecture on considering 3DGC tool was presented. It is based on three layers structure: XML-based representation for scene graph, Generic Binarization of XML content Specific Compression tools for 3D Graphics Primitives

250

The group acknowledged the advantages of such approach in promoting the AFX tools to the industry.

ResolutionRequest for subdivision of MPEG-4.Issue the first version of the WD.

75 Resolutions of 3DG

75.1 Output documents

75.1.1 The 3DG subgroup recommends to approve the following documents

No. Title TBP Available Editor14496-4 Conformance testing

9132 Text of ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

No 07/05/12 Mark Callow

9146 DoC on ISO/IEC 14496-4:2001/ FDAM16 (MPEG-J GFX Conformance)

No 07/05/12 Marius Preda

9133Text of ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

No 07/04/27 Jeong-Hwan Ahn

9147 DoC on ISO/IEC 14496-4:2001/ FPDAM21 (Geometry and Shadow Conformance)

No 07/05/04 Marius Preda

No. Title TBP Available Editor14496-5 Reference Software

9134Text of ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)PB: Mark has to clean up the code

N 07/05/12 Mark Callow

9148 DoC on ISO/IEC 14496-5:2001/ FDAM11 (MPEG-J GFX RefSoft)

N 07/05/12 Marius Preda

9135Text of ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)PB: Patrick has to send me the software from the CVS

N 07/05/04 Patrick Gioia

9149 Doc on ISO/IEC 14496-5:2001/ FPDAM13 (Geometry and Shadow RefSoft)

N 07/05/04 Marius Preda

No. Title TBP Available Editor14496-16 Animation Framework eXtension (AFX)

9136

WD 2.0 of ISO/IEC 14496-16:2006/AMD2 (Frame-based Animated Mesh Compression)

N 07/04/27 Marius Preda, Titus Zaharia

251

9137 WD 1.0 of ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

N 07/04/27 Patrick Gioia

9150 Request for ISO/IEC 14496-16:2006/AMD3 (3D MultiResolution Profile)

N 07/04/27 Marius Preda

9138 3D Graphics Core Experiments Description N 07/04/27 Khaled Mammou

9139 3D Graphics Compression FAQ 19.0 Y 07/05/12 Pierre Davy

No. Title TBP Available Editor14496-21 MPEG-J GFX

9140 Text of ISO/IEC 14496-21:2006/COR1 N 07/04/27 Mark Callow

No. Title TBP Available Editor14496-25 3D Graphics Compression Model

9141 Request for Subdivision of ISO/IEC 14496: Part 25 - 3D Graphics Compression Model

N 07/04/27 Marius Preda

9142 WD 1.0 for ISO/IEC 14496-25 Y 07/04/27 Marius Preda

75.2 Resolutions The 3DG subgroup recommends appointing Patrick Gioia (France Telecom) as the editor of ISO/IEC 14496-

16:2006/AMD3 and thanks him for taking the responsibility of this project. The 3DG subgroup initiates a new activity on applying MPEG 3D Graphics compression tools to third-parties

solutions for scene graph and graphics primitives’ representation and encourages external bodies to participate to this activity.

The 3DG subgroup recommends appointing Marius Preda (INT), Mark Callow (HI Corporation) and Jeong-Hwan Ahn (Samsung AIT) as the editors of ISO/IEC 14496-25 and thanks them for taking the responsibility of this project.

75.3 Establishment of 3DG Ad-Hoc Groups

N9143 AHG on 3DG documents, experiments and software maintenance

Mandate: 1. Maintain and edit 3DG documents2. Coordinate 3DG CE activity3. Coordinate 3DG related conformance and reference software

Chairmen: Jeong-Hwan Ahn (Samsung AIT),Ning Lu (Intel Corporation)

Duration: Until 81st Meeting Meetings Sunday before 81st meetingReflector: mpeg-3dgc AT gti. ssr. upm. Es

252

Subscribe: http://www.gti.ssr.upm.es/mailman/listinfo/mpeg-3dgc

76 Closing of the MeetingSee you in Lausanne.

253

Annex L – Test report

Source: Tobias Oelbaum, Chair

Report of Test meeting for the 80th MPEG meeting in San Jose, USA

77 Opening of the Meeting

Goals for the weekThe goals of this week are:

Refine the draft verification test plan for SVC, especially regarding test sequences and bit rates for the test

Provide input to JVT to the discussion on the MVC Deblocking Filter

78 Test Activities

Scalable Video Coding - Verification TestsThe Draft SVC Verification Test Plan has been updated. This especially includes the refinement of bit rates and test sequences that should be used for the test and a refinement of the single test scenarios. The test plan currently includes 4 scenarios for profile A and profile B and 2 scenarios for profile B Intra.Based on viewing sessions performed previous to the meeting at TUM in Munich bit rates for the proposed profile B were selected (related JVT-Contribution: JVT-V102). Sequences from JVT-W110 (related to profile B Intra) were viewed at the Meeting and at TUM. It was proposed to increase the bit rate for the test compared to the bit rates used in this contribution.Using the input from JVT-V102 and JVT-W110 four test sequences were identified that could be used for this test.Two late contributions to JVT (JVT-W131 and JVT-W135) were reviewed and related changes were made in the verification test plan. Two new sequences that were proposed by Layered Media for the use in the verification test were viewed. It was proposed to search for more challenging sequences. Layered Media will bring more sequences from the field of video conferencing to the next meeting.An AHG for preparing the verification tests has been set up.

Multi-view Video Coding – Deblocking FilterA short visual evaluation of JVT-W024 was conducted. The question was if the proposed extension of the deblocking filter for MVC would result in better subjective quality. Results of this evaluation (the subjective quality could be improved by this extension of the deblocking filter) were reported back to JVT.

254

79 Test Resolutions

Output Documents 8965 Draft SVC Verification Test Plan Version 3.0

AdHoc GroupsThe following AHG was set up:

N8993 AHG on SVC Verification Test

Mandate:1. To discuss test setups based on applications scenarios of the SVC

profiles2. To refine the verification test document3. To prepare the verification test

Chairman: Tobias Oelbaum (TU München, oelbaum@tum.de)Mathias Wien (RWTH Aachen, wien@ient.rwth-aachen.de)Associate Chairs:Vincent BottreauNathalie CammasAlex EleftheriadisJustin Ridge

Duration: Until 81st Meeting Meetings Yes (Sunday before the 81st Meeting)Reflector: mpeg-svt@lists.rwth-aachen.deSubscribe: To subscribe or unsubscribe, go to

http://mailman.rwth-aachen.de/mailman/listinfo/mpeg-svt

255

Annex M – ISG report

Source: ISG Chair, Marco Mattavelli (EPFL)

1 OverviewThe main work items of the Implementation Studies Subgroup in San Jose are:

1. The contributions to the Reconfigurable Video Coding (RVC) activity jointly with the video group for contribution review, review of results for the on going core experiments, editing of the RVC WD documents.

2. The review of the final core experiment results aiming at improving the finite precision DCT/IDCT specification selected at Hengzhou meeting considering possible further performance improvement and complexity reduction.

3. MPEG-4 Part 9 Reference HW description: The editing of the Study of the Third Edition of the TR The review of the new HDL module and associated documentation submitted for

integration in Part 9.

Input contributions to ISG group w.r.t. the above items are summarized according to the following table:

Input Contributions to ISG subgroup

M14276 Robert Turney (Xilinx) Marco Mattavelli (EPFL)

AHG report on MPEG-4 Part 9 Reference Hardware Description Phase 1 and 2”

M14434 Julien DuboisBarthelemy HeyrmanJohel Miteran et al.

Wildcard Platform Vs ML310

80 Detailed Report

80.1 The contribution to the activity on Reconfigurable Video Coding (RVC).

Most of the ISG time in San Jose has been spent in joint meetings with Video for the RVC subgroup work. The main issues of discussion were the evaluation of the results of the on-going core

256

experiments concerning the evolution and progress of the technology currently described in the WD. Major results reported are the:

limitations and bugs of the implementation of the MPEG-4 SP in terms of CAL FUs, the implementation of almost all FUs in CAL for AVC baseline new results of compression of DDL for a complete decoders no results were reported for the implementation of the flexile decoder based on BSDL

bitstream descriptions and transformations to CALML and CAL first proposals of methodologies for the conformance testing of RVC FUs studies and proposals for the efficient partitioning of FU for B-pictures, multiple reference

frames, intra prediction and for SVC, description of the RVC framework tool support and definition of future tool support.

All reviewed contributions are reported in the list below.

Contribution Number

Category Title

14301 MPEG-C RVC Functional Units naming process proposal

14340MPEG-B Compression of the RVC DDL Decoder Description

with BiM (results of Core Experiment 1.3 in RVC)

14374MPEG-C Functional units of inter-prediction under reasonable

system partition for RVC framework

14375 MPEG-C Conformance test tools of RVC functional units

14416MPEG-C Implementation of B frame support in RVC CAL

Model

14445 MPEG-B Core Experiment Result on CDDL14446 MPEG-B Proposed Text of RVC CE14447 MPEG-B Study on RVC Framework and Its Requirements

14448 MPEG-C Proposed text of the RVC FUs for MPEG-4 AVC (Results of CE 2.2)

14454MPEG-C Implementation of multiple reference frame support

in RVC CAL model

14457MPEG-C A scheme for implementing MPEG-4 SP codec in the

RVC framework

14463 MPEG-C Evolutions of RVC so as to handle SVC decoding

14473 MPEG-B Extension to support non-MPEG standards (ICT/ZJU) (Results of CE 1.6)

257

14474MPEG-B Exploration experiments of AVS decoder description

in RVC framework

14480MPEG-C Implementation of MPEG-4 AVC Deblocking Filter in

RVC CAL model

14490MPEG-B Reconfigurability potential of the MPEG-4 SP

decoder (results of CE 1.1)

14510 MPEG-C Proposal for adding ISO/IEC 23002-2 in RVC tool library

14542 MPEG-B/MPEG-C Liaison Statement to MPEG on RVC

14546 MPEG-B/MPEG-C

Description of tools for the RVC framework: editors simulator software and HDL code generators

80.2 Contributions on the specification of a finite precision IDCT

Several contributions have been received concerning cross check of core experiments and validation of results for finite precision IDCT performance and complexity. The main comments and major points of each contribution are reported in the table below.The more relevant results are reported by contribution M14506 in which it is shown how variant of the current CD algorithm (called Za) can achieve a further reduction of the implementation complexity for a negligible decrease of drift performances. This represents ~10% complexity savings compared to previous implementation (saves 2 shifts and 4 negations). Drift test shows negligible differences between the two. Another algorithm called L1m9 might be convenient for implementations because it can reuse blocks for (it can use 8 multiplier 26 additions and 8 shifts, but it does not pass the linearity test). The decision of the group was to move to Za algorithm and include it in the CD.

Core experiment reports:

14506 Yuriy Reznik

Summary of core experiments on fixed point IDCT/DCT

Report of precision results for 3 variants of the CD algorithm (Z0). Moving to Z0a could save 2 shifts. Drift test shows negligible differences between the two. Another algorithm L1m9 might be convenient to reuse blocks for implementation (it can use 8 multiplier 26 additions and 8 shifts, but it does not pass the linearity test.

14485 Zhibo NiLu Yu

IDCT Core Experiment Results

Experiments on variations of CD and other candidate

258

results. They are done with MPEG-2 and MPEG-4 including quarter-pel interpolation. Results do not show evidence for changing current CD algorithm.

14469

Honggang QiWen GaoDebin ZhaoSiwei Ma

Cross-check of IDCT core experiments

Results of 14485 have been cross-checked.

Summary: A variety of variations of the fixed-point IDCT specified in the CD have been successfully identified, with various trade-offs in regard to dynamic range, operation counts, operation types, etc. Drift analyses were performed for these IDCTs in H.263, MPEG-2, and MPEG-4 (with ½- and ¼-pel accurate MC).

Testbed updates:

14346 Yuriy ReznikUpdated 23002-1 IDCT precision testbed

Testbed update

14347 Yuriy ReznikArianne Hinds

Updated H.263-based IDCT testbed

Row column implementation according to previous standards.

14348 Arianne T. Hinds

Updated MPEG-4 IDCT Testbed

Updates including MPEG-4 row-column implementation.

14379 Arianne T. Hinds

Updated T.83 testbed for IDCTs

Conformance test for JPEG update for row-column implementation

14380 Zhibo NiUpdated MPEG-2 IDCT Testbed

Inclusion of row first implementation and

14403 Arianne T. Hinds

Updated TM5 MPEG-2 Testbed

Addition of H.263 and TM5 with row fist implementation

Summary for all contributions: all testbeds have been updated with all modifications included in the approved CD including row-column Implementations. Testbeds have also been updated to include existing fixed point IDCT algorithms from MPEG-2 TM5, H.263 and XVID.

259

Editing reports:

14310

Yuriy A. ReznikGary SullivanArianne T. Hinds

Study Text of ISO/IEC 23002 CD (editors input)

Change of title adding implementation and taking out “transform”. Definition of the transform is changed. Editing according to NB comments received at Marrakech meeting.

14311 Yuriy Reznik

Study Text of ISO/IEC 23002-1/PDAM1 (editors input)

Amendments of reference SW. Software overview. Mainly cleanup of previous text without any relevant change.

Summary: A variety of editorial issues were identified with the current CD text and improvements were proposed to address them.

Conformance tests:

14531 Arianne T. Hinds

Fixed-Point IDCT Conformance Tests

Report of conformance tests for the CD algorithm. Also other algorithms pass conformance tests.

14509 Yuriy Reznik

Cross-check of IDCT conformance tests

Cross check of the results is confirming the results of 14531.

Summary: A testbed was provided for verification of CE IDCTs using the methodology of MPEG-2 video conformance testing. The results were provided and cross-checked.

Drift phenomena analysis and studies:

14544 Zhibo NiLu Yu

On the Problem of Quarter Pixel Motion Compensation

Results showing severe drift results in case of quarter pel interpolation for MPEG-4 ASP. An analysis of the reasons of such drift problems is provided. The contribution presents striking evidence of the need of bit exact match between encoder and decoder IDCT implementations.

Summary: Analysis of drift propagation with 1/4-pel MC in MPEG-4 P2. This analysis explains empirically observed phenomena of drift propagation with ¼-pel MC.

260

Contributions on IDCT design:

14359

Yi-Shin TungChung Hsuan KuoMing Chung HsuJa-Ling Wu

Consider Row-Transform-First IDCT in 23002-2 and the Fixes to 23002-2 CD

The contribution presents the implementation efficiency reasons for which implementing 2-D IDCT where 1-D row IDCT are processed first and then columns are processed after is advantageous. This suggestion has already been accepted and included in the study text of the FCD.

Summary: Arguments provided in support of implementing 2D IDCTs with 1D row- processing first, followed by the column- processing.

80.3 The progress in the development of the MPEG-4 “Part 9 Reference Hardware Description”

The ISG activity at the San Jose meeting has mainly been devoted to the review of the received contribution (M14434), the editorial work for third edition of the technical report,

81 ResolutionsThe above activities have led to the following resolutions and output document approval.

82 Resolutions related to MPEG-4

Part 9 Reference Hardware Description

The ISG subgroup recommends to approve the following documents

No. Title TBP Available14496-9 Reference Hardware Description

8994 Status of HDL submissions and commitments for MPEG No 07/04/278995 Study of ISO/IEC DTR 14496-9 No 07/04/27

83 Resolutions related to MPEG-B

261

Part 4 Codec Configuration Representation

The video subgroup and the ISG recommend to approve the following documents

No. Title TBP Available23001-4 Codec Configuration Representation

8979 WD 4 of ISO/IEC 23001-4 No 07/05/04

MPEG notes that the RVC project is about developing a full collection of individual coding tools organized in the video tool library and a generic framework that can be used to make an implementation of any MPEG video coding standard. Further MPEG recognises the benefit of having the framework be capable of additionally supporting the implementation of video coding standards from other organizations with which a collaboration can be established. As part of this project, an identification mechanism will be developed whereby MPEG video coding tools will be identified by MPEG and video coding tools from other organizations can be identified via a registration authority.

The video subgroup thanks AVS for their liaison and for providing the specification and reference software of their standard as needed for the development of the capability of ISO/IEC 23001-4 to support non-MPEG toolboxes.

MPEG invites organisations who would like to collaborate in the development of the framework to join MPEG in making the framework support all widely deployed video codecs.

84 Resolutions related to MPEG-C

Part 2 Fixed point 8x8 DCT/IDCT

The ISG and the video subgroups recommend changing the title of 23002-2 to “Fixed-point 8x8 IDCT and DCT”

The ISG and the video subgroups recommend to approve the following documents

No. Title TBP Available23002-2 Fixed point 8x8 DCT/IDCT

8982 Disposition of Comments on ISO/IEC CD 23002-2 No 07/04/278983 Text of ISO/IEC FCD 23002-2 Fixed-point 8x8 IDCT and DCT No 07/05/04

The video subgroup thanks the National Bodies of Germany and US for their valuable ballot comments on ISO/IEC CD 23002-2.

262

Part 4 Video Tool Library

The ISG and the video subgroups recommend to approve the following documents

No. Title TBP Available23002-4 Video Tool Library

8984 WD 4 of ISO/IEC 23002-4 No 07/05/258985 Description of Core Experiments in RVC No 07/05/048986 RVC Simulation Model (RSM) V4.0 No 07/05/258987 RVC Work Plan No 07/05/048988 RVC Conformance Testing Working Draft 1.0 No 07/05/148989 Description of Exploration Experiments for Toolbox Extensions No 07/05/14

263

Annex N – Liaison report

Source: Kate Grant, Chair

The Liaison group received the following input documents and discussed them at their meeting on Tuesday April 24th:

No. TitleLiaison Statements

14285 Liaison Statement from W3C (MMSEM)Information on current W3C MMSEM work: in particular links to 2 documents: Image Annotation on the Semantic Web and Multimedia Semantics on the Web: Vocabularies14297 Liaison Statement from 3GPPInput on LASeR from 3GPP SA4 group14300 Liaison Statement from ITU-T FG IPTVEnclose FG IPTV-R-0021: Report of the 3rd Focus Group on IP Television (IPTV) meeting14305 Liaison Statement from DVD ForumConcern regarding backward compatibility problems with N 8859 MPEG-2 Systems DCOR14313 Liaison Statement from IEC TC100Text of CDV of Edition 2 of IEC 61937-3 (currently under ballot) for information14314 Liaison Statement from IEC TC100Text of CDV of IEC 61966-2-5 (opRGB) (currently under ballot) for information14331 Liaison Statement from ETSIUpdate on issues regarding proposed optional use of MPEG-4 ER AAC-LD for NG-DECT superwideband conversational applications.14342 Liaison Statement from CEACEA IPTV Roadmap and Phase 2 Report provided for comment before 15th June14349 Liaison Statement from SMPTE Concern regarding backward compatibility problems with N 8859 MPEG-2 Systems DCOR14353 Liaison Statement from ATIS IIFIPTV Interoperability Specification for the IIF Default Scrambling Algorithm (ATIS-0800006) provided for the information and comment14354 Liaison Statement from ITU-T SG16 (Q10/16)Selected a reference codec for ITU-T G.722.1 fullband extension standardization that is publicly available (LAME MP3, http://lame.sourceforge.net).14362 Liaison Statement from DVBRequest MPEG-7 schemas made available online for automatic retrieval 14413 Liaison Statement from TTA Information on growth of market in Korea and need for rapid progression of DMB MAF 14533 Liaison Statement from 3D ConsortiumInformation about consortium and requirement for FTV standardisation14534 Liaison Statement from TC46/SC9/WG7Nominating liaison representative and providing background information

264

14535 Liaison Statement from JCPInformation that comments from 79th meeting reflected in current version of JSR-28714541 Liaison Statement from AVSProviding AVS specification and reference software to assist collaboration between MPEG and AVS on RVC and work on identifying general-purpose common elements14547 Liaison Statement from AESProject AES-X159, Carriage of PCM with MPEG Surround data over AES3 initiated in SC-02-0214548 Liaison Statement from FLOForumInformation on use of AVC in MediaFlo and that work on Rich Media is ongoing

The Liaison group prepared the following output documents:

No. TitleLiaison Statements

8919 Liaison Statement to WG1Provide information on MPEG-7 Query Format work and CD text for comment8920 Liaison Statement to IETFProvide information on new mime type8921 Liaison Statement to KhronosProvide information on new work on 3D Graphics Compression Model, and invite input8922 Liaison Statement to ISO TC184 SC4Provide information on new work on 3D Graphics Compression Model, and invite input8923 Liaison Statement to 3GPPProvide detailed information relating to LASeR8924 Liaison Statement to W3CProvide information on Photo Player, an implementation for digital photo libraries8925 Liaison Statement to ITU-T FG/IPTV concerning M3WUpdate on status of M3W standardisation8926 Liaison Statement to ITU-T FG IPTVStudying documents in work on identifying IPTV requirements. Provide information on MAFs and attach FCD of Media Streaming Player8927 Liaison Statement to SMPTE Text of revised DCOR on MPEG-2 systems (which addresses their concerns) for comment8928 Liaison Statement to DVD ForumText of revised DCOR on MPEG-2 systems (which addresses their concerns) for comment8929 Liaison Statement to ETSI Response to incoming liaison, offering to provide further information if required8930 Liaison Statement to SMPTE re file format Provide document on TuC for ISO base media file format for comment8931 Liaison Statement to DVB MPEG-7 schemas to be made available on line at ITTF web site8932 Liaison Statement to JCPAppreciation that updated JSR-287 specification includes comments from 79th meeting8933 Liaison Statement to CEA

265

Information on MPEG specifications which relate to the issues being studied (DRM, QoS etc) and information on MAFs8934 Liaison Statement to ATISInformation on MPEG specifications which relate to the issues being studied8935 Liaison Statement to SMPTE re RVCInvite experts to participate in development of RVC. Information on AVS collaboration. 8936 Liaison Statement to 3D ConsortiumThank them for information, inform them of start of work on FTV 8937 Liaison Statement to FLOForumThank them for information, update them on progress of SVC standardisation8938 Liaison Statement to TC46/SC9/WG7Welcome liaison representative, send them MPEG document on URNs ( for comment8939 Liaison Statement to AVSThank them for providing AVS specification and reference software for RVC development and welcome collaboration on development of RVC framework.Other Documents8940 Response to National Bodies Responses to USNB and Italian NB8941 List of Organisations with which MPEG entertains liaisons (as of April 2007)Updated with latest information

266