Presentation is loading. Please wait.

Presentation is loading. Please wait.

4EyesFace-Realtime face detection, tracking, alignment and recognition Changbo Hu, Rogerio Feris and Matthew Turk.

Similar presentations


Presentation on theme: "4EyesFace-Realtime face detection, tracking, alignment and recognition Changbo Hu, Rogerio Feris and Matthew Turk."— Presentation transcript:

1 4EyesFace-Realtime face detection, tracking, alignment and recognition Changbo Hu, Rogerio Feris and Matthew Turk

2 Overview  Introduction  Face Detection and Pose tracking  Face Alignment  Face Recognition  Conclusions

3 Introduction DetectionPose trackingAlignmentRecognition

4 Introduction  Why this is a difficult problem? Facial Expressions, Illumination Changes, Pose, etc.  Object Develop a fully automatic system, suitable for real-time applications to locate and track human faces, then to align and recognize the face. Evaluate it on a large dataset.

5 Face Detection [Viola and Jones, 2001]  Simple features, which can be computed very fast.  A variant of Adaboost is used both to select the features and to train the classifier.  Classifiers are combined in a “cascade” which allows background regions of the image to be quickly discarded.

6 Face detection

7 Pose tracking Based on Kentaro Toyama ’ s IFA framework

8 Face Alignment  Active Appearance Model (AAM) Statistical Shape Model (PCA) Statistical Texture Model (PCA)

9 Face alignment  Problem: Partial Occlusion  Active Wavelet Networks (AWN) (on BMVC ’ 03) Main idea: Replace AAM texture model by a wavelet network

10 Face Alignment Similar performance to AAM in images under normal conditions. More robust against partial occlusions.

11 Face Alignment Using 9 wavelets, the system requires only 3 ms per iteration. In general, at most 10 iterations are sufficiently for good convergence (PIV 1.6Ghz).

12 Multi-View Face Alignment  View selection by pose tracker

13 Multi-View Face Alignment

14 Face recognition  online recognition HMM based face recognition

15 Face recognition  Large dataset evaluation  FERET DataSet  1196 different individuals  With ground truth of eye corners

16 Face recognition

17

18 Face Recognition

19

20 Conclusion  We develop a system to do human face detection, tracking, alignment and recognition  In this system, we invented new methods AWN and extent to multi- view AWN  We implement the related detection and pose tracking  Evaluate our method on large dataset


Download ppt "4EyesFace-Realtime face detection, tracking, alignment and recognition Changbo Hu, Rogerio Feris and Matthew Turk."

Similar presentations


Ads by Google