Dominant singular value decomposition representation for face recognition

We propose in this paper a new dominant singular value decomposition representation (DSVDR) method for face recognition. Motivated by the fact that each grayscale face image can be decomposed into a composition of a set of bases by the well-known singular value decomposition (SVD) technique and each basis contains different discriminative and reconstructive information for face representation, we present a new face representation method to select a subset of important bases and regulate their singular values (SVs) according to their discriminative and reconstructive power simultaneously for face recognition. Experimental results demonstrate the efficacy of the proposed method.