Facial Recognition and Verification using Gabor Wavelets and Kernel Methods (TA-P2)

Author(s) : Li Bai (Nottingham University, UK) Linlin Shen (University of Nottingham, UK) Phil Picton (University College Northampton, UK)

Abstract : A novel Gabor-Kernel face recognition method is proposed in this paper. This involves convolving a face image with a series of Gabor kernels at different scales, locations, and orientations to obtain feature vectors. Kernel methods such as Kernel Principal Component Analysis (KPCA) and Kernel Discriminant Analysis (KDA) are then applied to the feature vectors for dimension reduction as well as class separability enhancement. A database of 600 frontal-view face images from the FERET face database is used to test the methods. Experimental results demonstrate the advantage of Kernel methods over classical Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). Improvements are also observed when Gabor features instead of the original images are used. The method achieves 92% recognition accuracy using only 35 features of a face image. Following is a brief summary of each section of the paper.

Menu