View Invariant Face Recognition from 2D and 3D Images

Y. Wang, C.-S. Chua, and Y. Ren (Singapore)

Keywords

View invariant, Gaborilter responses, Point Signature, Hausdorff Distance

Abstract

This paper presents a view invariant face recognition system based on both 3D range data as well as 2D gray-level facial images. An irregular 2D mesh labeled by twelve landmarks and a 3D region labeled by four landmarks are deined for each model for feature extraction. Nodes of the 2D mesh are described by Gaborilter responses and 3D points are represented by Point Signature. Each person provides a frontal model face and discriminating features, are automatically selected to build the model library in the 2D domain. To classify test faces under varying poses, a robust Hausdorff Distance is used to handle the possible case of matching incomplete data. The best matched model is determined based on the linear integration of matching results in 2D and 3D domains. Experimental results based on our database involving 80 per sons with different facial expressions and viewpoints and the ORL face database have demonstrated the promising performance of our algorithm.

Important Links:



Go Back


IASTED
Rotating Call For Paper Image