An Improved Algorithm of Facial Expression Recognition by Fusion of Sparse Representation

Abstract

Continuing effects are being made to further increase the accuracy of facial expression recognition due to the fact that at present the expression recognition rate is relatively low for practical use. The algorithm, which is based on fusion of sparse representation and local binary patterns (LBP), gets good recognition rate on Japanese Female Facial Expression database JAFFE in [1]. In this paper, we propose an algorithm which is a modification of the algorithm proposed in paper [1]. The new algorithm first solves the sparse representations both on raw gray facial expression images and two group LBP features of these images. Then we can obtain the two expression recognition results on both sparse representations. Finally, the final expression recognition is performed by fusion on the two results. The experiment results on Japanese Female Facial Expression (JAFFE) database demonstrate that the new algorithm is better than the original fusion algorithm.