Transcript
Page 1: Synthetic animation of deaf signing

1

Synthetic animation Synthetic animation ofof

deaf signingdeaf signingRichard KennawayRichard Kennaway

University of East AngliaUniversity of East Anglia

Page 2: Synthetic animation of deaf signing

2

VISICASTAutomated deaf signing of

broadcast television

• Translation of text to sign

• Animation of signs

• Transmission over broadcast

channel

• Rendering of avatar

Page 3: Synthetic animation of deaf signing

3

Motion capture

• Very lifelike animation

• Time-consuming to set up• Blending of signs• Combining signs from different signers

Page 4: Synthetic animation of deaf signing

4

Hand-crafted animation

• Can give good animation

• Time-consuming (1/2 hour per sign)• Blending of signs still required

Page 5: Synthetic animation of deaf signing

5

Synthesis from semi-abstract transcription

• Quick to create lexicon (a few minutes to transcribe a sign)

• Instantly retargettable to any avatar with humanoid topology

• Automatic blending

• Low bandwidth

Page 6: Synthetic animation of deaf signing

6

HamNoSys: Hamburg Notation System

DGS “GO-TO (by car)”

Page 7: Synthetic animation of deaf signing

7

SiGML: an XML-isation of HamNoSys

<sign_manual both="true"><handconfig extfidir="uo" palmor="l"

handshape="finger2" thumbpos="open"/><tgt_motion>

<directed_motion direction="o" curve="u"/><handconfig extfidir="do"/>

</tgt_motion></sign_manual>

Page 8: Synthetic animation of deaf signing

8

Animation of HamNoSys

Make explicit everything HamNoSysleaves implicit or fuzzy:

• position• elbows and shoulders• speed• trajectories

Page 9: Synthetic animation of deaf signing

9

Naturalistic animation

A hard problem in general (e.g. walking).

Easier for signing: no interaction with environment, ignore gravity.

Page 10: Synthetic animation of deaf signing

10

desiredposition

k1

actualposition desired

velocity

error

gain

actualvelocity

gain

acceleration

error

k2

Page 11: Synthetic animation of deaf signing

11

desiredposition

k1

actualposition desired

velocity

error

gain

Simulated physics

A simple linear control system

actualvelocity

gain

force/torque

error

k2

Page 12: Synthetic animation of deaf signing

12

Controller response

Page 13: Synthetic animation of deaf signing

13

Inverse kinematics

Hand position and orientation given by HamNoSys

From these, compute joint angles from clavicle to wrist (inverse kinematics).

3 degrees of freedom per arm left undetermined:respect the limits of the jointsavoid the arm passing through the body

Page 14: Synthetic animation of deaf signing

14

Stick-figure avatars

Useful for developing animations:

• easier to render, so more frames per second

• skeleton gives clearer view of motion

• prototyping tool only, not intended for end user!

Page 15: Synthetic animation of deaf signing

15

VRML for prototyping

Textual description language for 3D animated scenes.

H-Anim standard for articulated humanoid figures.

H-Anim incorporated into MPEG-4.

Virtual Reality Modelling Language

Page 16: Synthetic animation of deaf signing

16

Ambient motion

Mix synthetic animation with motion-captured “ambient motion” for the spine and head.

If only arms, hands, and face are animated, the result is stiff.

Page 17: Synthetic animation of deaf signing

17

Next steps

Implement the whole of SiGML/HamNoSys

Facial animation:blend motion-capture dataimplement HamnoSys 4

Tests


Top Related