Compound local binary pattern (CLBP) for robust facial expression recognition

The local binary pattern (LBP) operator has been proved to be a simple and effective approach for facial feature representation. However, the LBP operator thresholds P neighbors at the value of the center pixel in a local neighborhood and encodes only the signs of the differences between the gray values. Thus, the LBP operator discards some important texture information. This paper presents a new local texture operator, the compound local binary pattern (CLBP), and a feature representation method based on CLBP codes for facial expression recognition. The CLBP operator combines extra P bits with the original LBP code, which are used to express the magnitude information of the differences between the center and the neighbor gray values. We empirically evaluate the effectiveness of the proposed feature representation for person-independent expression analysis. Extensive experiments show the superiority of the CLBP method against some other appearance-based feature representation methods.