Fusion of Genetic Algorithm with Tensor based Algorithms for Face Recognition

This paper proposes a new algorithm based on multi PCA approach and tensor based algorithm. Using MPCA and DATER, the features will be extracted and applying aGenetic Algorithm, the best eigenvectors have been chosen for the next step of face recognition. In this method, some smalleigenvectors also will be used for dimension reduction. Our approach brings two advantages. First, optimal bases for dimensionality reduction could be derived from GA-MPCA. Second, the computational efficiency of DATER is improved. The resulting algorithm is more successful (in terms of recognition rate) than the common Eigenfaces algorithm. Its effectiveness is proved for two standard databases (ORL and FERET databases), which includes different modes and poses, like illuminations, expressions and with or without glass or scarf

Keywords:

Genetic Algorithm , Tensor objects , Multilinear discriminant analysis , subspace learning , multilinear principal component analysis