Multiplication fusion of sparse and collaborative representation for robust face recognition

摘要: Sparse representation based classification (SRC) and collaborative (CRC) are two well-known methods in representation-based for face recognition. SRC emphasizes on the role of sparsity coding, while CRC obtains an enhanced sparse with contributions from all classes. Previous studies show that coefficients is significant robust may amplify nonlocal contribution other However, parameters closely associated it very difficult to solve optimal solutions. This paper proposed a novel image method integrates by multiplication operation solutions obtain representation. We conducted experiments several benchmark databases experimental results showed our algorithm produces higher accuracy than both CRC.