Face gesture recognition based on clustering algorithm

In recent years, face recognition has been increasingly applied to all aspects of social life. Face recognition actually includes a series of related technologies for constructing face recognition systems, including face image acquisition, face location, face recognition preprocessing, identity confirmation, identity search, etc. However, most face recognition technologies today have high requirements on the pose of the face, and the robustness to the face pose is very low. Therefore, multi-pose face recognition has become a hot trend in face recognition. This paper uses the ASEF human eye localization algorithm to locate the human eye by preprocessing in the Multipie database. Then, the clustering algorithm is used to cluster the face pose to achieve the effect of gesture recognition.