An Associate-Predict Model for Face Recognition

1. An Associate-Predict Model for Face Recognition CVPR 2011 Qi Yin1,3Xiaoou Tang1,2Jian Sun3 1. Department of Information Engineering The Chinese University of Hong Kong 2. Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences, China 3. Microsoft Research Asia

2. Outline • Introduction • Identity Data Set and Face Representation • Associate-Predict model • Switching Mechanism • Experimental Results

3. Introduction • Appearance-based for face recognition • Inevitable obstacle • Associate-Predict model • The studies of brain theories

4. Introduction

5. Identity Data Set and Face Representation • Identity data set • Face representation

6. Identity data set • 200 identities from the Multi-PIE data set • 7 pose • 4 illumination

7. Face representation • Representation at the facial component level • 12 facial components • Face F = (f1, f2, ..., f12) • fi for each component

8. Associate-Predict model • Appearance-prediction model • Likelihood-prediction model

10. Appearance-predictionmodel • Two input faces • Setting : SA , SB • A and B are facial components • Select the specific face image setting is equal to SB • component A’ from this image

11. Appearance-predictionmodel • dA = |fA' − fB| • distance between the components • dB= |fB' − fA| • Final distance between A and B:1/2 (dA + dB)

12. Appearance-predictionmodel • Adaptive distance dp • αA and αB : weight • After the “appearance-prediction” on all 12 facial components , we can obtain a new composite face

13. Likelihood-prediction model • Using classifier measure the likelihood of B belonging to A • Positive training samples • Input face • the K most alike generic identities

15. Switching Mechanism • Implement this switching mechanism • facial components : A and B • settings : SA = { PA , LA } and SB = { PB , LB } • Categorize the input pair into two classes • “comparable” • “not comparable” • based on the difference of SAand SB

16. Switching Mechanism • Comparable class • {|PA − PB| < 3 } and {|LA − LB| < 3 } • Not comparable class • the rest situations

17. Switching Mechanism • The final matching distance dsw • da: the direct appearance matching • dp: the associate-predict model

18. Experimental Results • Experiments on the Multi-PIE and LFW data sets • Basic comparisons • Results on benchmarks

19. Basic comparisons • Holistic vs. Component

20. Basic comparisons • Positive sample size • number of positive samples is 1 + 28*k • “1” is the input sample • K is the selected number of top-alike associated identities

21. Basic comparisons • K = 3 as the default parameter

22. Basic comparisons • Switching mechanism • the switch model can effectively improve the results on both benchmark

23. Results on benchmarks • Multi-PIE benchmark • LFW benchmark

24. Multi-PIE Benchmark

25. LFW benchmark