1. facial expression editing in video using a temporally- smooth factorization 2. face swapping:...

24
1. Facial Expression Editing in Video Using a Temporally-Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Upload: jamison-peirson

Post on 16-Dec-2015

216 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

1. Facial Expression Editing in Video Using a Temporally-Smooth Factorization

2. Face Swapping: Automatically Replacing Faces in Photographs

Page 2: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Facial Expression Editing in Video Using a Temporally-Smooth Factorization

Fei Yang, Lubomir Bourdev, Eli Shechtman, Jue Wang, Dimitris Metaxas

CVPR 2012

Page 3: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Goal

The goal is to allow for semantic-level editing of

expressions in a video:

magnifying an expression

suppressing an expression

replacing by another expressions

3

Page 4: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Example

4

Page 5: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Challenges

Natural expression

Different parts changes accordingly

Unique identity

Temporal coherency

5

Page 6: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Related Work 2D based methods

[Theobald09], [Liu01], [Williams90], …

3D based methods [Blanz03], [Pighin98], …

Expression flow [Yang11]…

Frame reorder method [Bregler98], [Kemelmacher- Shlizerman11]

Tensor factorization methods [Vlasic05], [Dale11]…

6

Page 7: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Algorithm

7

Expression Information

Identity Information

3D Tensor Model - [Vlasic et al siggraph05]

Modify

Page 8: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Mode-n Product

8

Page 9: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Algorithm

goal to identify a and

method 2D v.s. 3D

frame t

Minimize: | – |

=

Weak Projective Matrix Rt

Page 10: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Algorithm

Fitting Error:

Shape Distribution Constraint:

Temporal coherence:

Page 11: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Algorithm

11

Levenberg-Marquardt (Siggraph98)

Page 12: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Algorithm

12

Adjust to achieve expression

modification Dynamic Time Warping (DTW)

[Sakoe78]

Residual Expression Flow

Correcting boundary compatibility

Page 13: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Results

13

Page 14: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Face Swapping: Automatically Replacing Faces in Photographs

Dmitri Bitouk Neeraj Kumar Samreen Dhillon Peter Belhumeur Shree K. Nayar

Siggraph 2008

Page 15: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Examples

15

Page 16: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Goals

16

For an input image:

Automatically find the best candidate

Automatically replace the face

Automatically color and lighting adjustmet

Page 17: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Library Building

17

OKAO face detector to detect face pose [Omron07]

Page 18: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Process

18

Page 19: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Alignment

19

Pose, Resolution, and Image Blur:

Yaw, pitch threshold between two images ( )

Eye distance as a measure of distance (80%)

Similarity of the blur degrees [Kundur and

Hatzinakos 1996; Fergus et al. 2006]

Page 20: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Color and Lighting

20

To ensure the similarity between the replaced and original face, a linear combination of 9 spherical harmonics [Ramamoorthi and Hanrahan 2001; Basri and Jacobs 2003] is used as measure metric:Each pixel I(x, y) can be approximated by:

Distance:

Page 21: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Seam Signature

21

256-by-256 patch from the face is used for replacement.

Unfold:

L2 Norm is used to compute the distance

Page 22: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Appearance Adjustment

Using simple scaling on the Harmonics coefficients

, are the original and replacement images

Scale the replaced image

Page 23: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

Results

Page 24: 1. Facial Expression Editing in Video Using a Temporally- Smooth Factorization 2. Face Swapping: Automatically Replacing Faces in Photographs

The End

24

Any Questions

?