Image-Set Based Collaborative Representation for Face Recognition in Videos

摘要

Video-based face recognition has become one of the hottest topics in the domain of face recognition because it has a wide range of applications in multi-media processing conference, human-computer interaction, judicature identification, video surveillance, and entrance controlling, etc. Methods of video based face recognition could be divided to be two main aspects: the models used to represent the individual image sets; and the similarity metric used to compare the models. Based on image-set based object classification methods, we present an image-set based on collaborative representation based classification (SCRC) method for face recognition in videos. Firstly, the query face video is divided to be many sub sets. Secondly, every sub set is represented by the collaborative representation based classification. Finally, we combine the recognition results of sub sets to be a final classification. Experiments test on three public video face datasets, the experimental results demonstrate that the proposed SCRC method can be able to outperform a number of existing state-of-the-art ones.