Statistical analysis of Gabor-filter representation

A successful face recognition system calculates similarity of face images based on the activation of multiscale and multiorientation Gabor kernels, but without utilizing any statistical properties of that representation. A method has been developed to weight the contribution of each element (1920 kernels) in the representation according to their power of predicting similarity of faces. The same statistical method has also been used to assess how changes in orientation (horizontal and vertical), expression, illumination and background contribute to the overall variance in the kernel activations. Weighting the elements in the representation according to their discriminative power has shown to increase recognition performance on a Caucasian and on a Japanese test image-set. It has also been demonstrated that such weighting method is particularly useful when data compression is a key requirement.

[1]  D. Watt.  Visual Processing: Computational Psychophysical and Cognitive Research , 1990 .

[2]  Hyeonjoon Moon,et al.  The FERET evaluation methodology for face-recognition algorithms , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[3]  Joachim M. Buhmann,et al.  Distortion Invariant Object Recognition in the Dynamic Link Architecture , 1993, IEEE Trans. Computers.

[4]  Hartmut Neven,et al.  PersonSpotter-fast and robust system for human detection, tracking and recognition , 1998, Proceedings Third IEEE International Conference on Automatic Face and Gesture Recognition.

[5]  Vicki Bruce,et al.  Processing Images of Faces , 1992 .

[6]  M. Young,et al.  Sparse population coding of faces in the inferotemporal cortex. , 1992, Science.

[7]  Hartmut Neven,et al.  Automatic pose estimation system for human faces based on bunch graph matching technology , 1998, Proceedings Third IEEE International Conference on Automatic Face and Gesture Recognition.