Shannon Wavelet Kernel based Subspace LDA Approach in Face Recognition
Wen-Sheng Chen, Bin Fang, Pong Chi Yuen, Jianhuang Lai · 2006
It is well-known that the distribution of face images with different pose, illumination and face expression is complex and nonlinear. The traditional linear methods, such as linear discriminant analysis (LDA), will not give a satisfactory performance. In addition, LDA always suffers from small sample size (S3) problem, which always occurs when the sample size is smaller than the dimensionality of feature vector. To overcome these limitations, Shannon wavelet kernel combining with subspace LDA (SWKSLDA) algorithm is developed. Two databases, namely FERET and CMU PIE databases, are selected for evaluation. Comparing with the existing LDA-based methods, the proposed method gives superior results