Automatic Classification Video for Person Indexing

Person retrieval and indexing in video sequences is a challenging task for many multimedia applications. This paper proposes a new method that index the person based on the similarity. Firstly, the persons in a shot are detected and tracked through face detector and continuously adaptive mean shift algorithm. Then mid-level features such as clothes colors and voice are applied to represent the person. An unsupervised cluster method is performed to cluster the person for further indexing. At last, the cluster is validated and refined by the voice feature. Experimental results of proposed method are presented, and the method has been found to be effective.