Gabor-Kernel Fisher analysis for face recognition

Abstract

Kernel based methods have been of wide concern in the field of machine learning. This paper proposes a novel Gabor-Kernel Fisher analysis method (G-EKFM) for face recognition, which applies Enhanced Kernel Fisher Model (EKFM) on Gaborfaces derived from Gabor wavelet representation of face images. We show that the EKFM outperforms the Generalized Kernel Fisher Analysis (GKFD) model. The performance of G-EKFM is evaluated on a subset of FERET database and CAS-PEAL database by comparing with various face recognition schemes, such as Eigenface, GKFA, Image-based EKFM, Gabor-based GKFA, and so on. © Springer-Verlag Berlin Heidelberg 2004.