Entropic-GWT based feature extraction and LBPSO based feature selection for enhanced face recognition

The appearance of the face will vary drastically when pose, illumination, background and expression change. Variations in these conditions make Face Recognition (FR) an even more challenging and difficult task. In this paper, we propose two novel techniques, viz., Entropic Gabor Wavelet Transform (Entropic-GWT) and Logarithmic Binary Particle Swarm Optimization (LBPSO), to improve the performance of a FR system. Entropic-GWT is a feature extraction technique used to minimize the dimensionality of the Gabor feature vector. LBPSO is a feature selection evolutionary algorithm which is used to search the feature space for a global optima. Experimental results show the promising performance of the proposed techniques for FR on three benchmark Face databases, namely, Color FERET, CMU PIE, and Extended YaleB.