Regularized Weighted Collaborative Representation with Maximum Likelihood Estimation for Facial Expression Recognition

Sparse representation is a new approach that has received significant attention for image classification and recognition. This paper presents a regularized weighted collaborative representation based classification algorithm with maximum likelihood estimation. In this method, we modify the collaborative representation fidelity term by introducing regularization factor. On the other hand, by maximizing the likelihood of collaborative fidelity term, the proposed method minimizes the collaborative residuals and improves the effectiveness of facial expression recognition system effectively. The experimental results on the Cohn-Kanade database demonstrate that the proposed method is effective in classification and can achieve a satisfactory recognition result.