Probabilistic Graphical Model for Auto-Annotation, Content-Based Retrieval, and Classification of TV Clips Containing Audio, Video, and Text | IEEE Conference Publication | IEEE Xplore