Traditional studies on micro-expression feature extraction primarily focused on global face from all frames. To improvethe efficiency of feature extraction, this paper proposes a new framework based on the local region and the key frame torepresent facial micro-expressions. Firstly, the face feature point detection technique is used to acquire the coordinates ofthe 68 key points, and the region of interest is divided by those key point coordinates and the action unit. Secondly, inorder to remove redundant information in the micro-expression video sequence, structural similarity index (SSIM) isused to select key frames for each local region of interest. Finally, the dual-cross patterns (DCP) are extracted for thelocal regions of interest and are concatenated into a feature vector for the final classification. The experimental resultsshow that compared with the traditional micro-expression method, the proposed method has higher recognition rate andachieves better time computation performance.
展开▼