Kernel Collaborative Representation with Regularized Least Square for Face Recognition

Zhenyu Wang, Wankou Yang, Jun Yin, Changyin Sun

Abstract: Sparse representation based classification (SRC) has received much attention in computer vision and pattern recognition. SRC is very slow since it needs optimize an objective function with L1-Norm. SRC consists of two parts: collaborative representation and L1-norm constrain. Based on SRC, collaborative representation based classification with regularized least square (CRC_RLS) is prosed. CRC_RLS is a linear method in nature. There are many variations of illumination, expression and gesture in face images. So face recognition is a nonlinear case. Here we propose a kernel collaborative representation based classification with regularized least square (Kernel CRC_RLS, KCRC_RLS) by implicitly mapping the sample into high-dimensional space via kernel tricks. The experimental results on FERET face database demonstrate that Kernel CRC_RLS is effective in classification, leading to promising performance.