Séminaire MIDI : Son Vu

Titre du séminaire et orateur

Contributions to unconstrained face analysis and recognition.
Son Vu.

Date et lieu

Mardi 16 avril 2013, 15h.
ENSEA, salle 384.


This talk presents recent contributions to unconstrained face analysis and recognition with respect to many stages in a facial biometrics system.  An illumination normalization method simulating the performance of human retina is proposed as preprocessing algorithm. A series of efficient descriptors exploiting the gradient information for face representation is presented. Those descriptors are discriminative and robust to exterior variations (variations of pose, illumination, expression and even image quality). Also, statistical models for robust face recognition across poses are presented. Several different applications of those methods such as texture classification, region description and face anti-spoofing are also discussed. All those methods producing the state-of-the-art performance are very fast and are therefore suitable for the constraints upon real-time of surveillance applications.