Semantic Topic Discovery for Lecture Video

With more and more lecture, videos are available on the Internet, on-line learning and e-learning are getting increasing concerns because of many advantages such as high degree of interactivity. The semantic content discovery for lecture video is a key problem. In this paper, we propose a Multi-modal LDA model, which discovers the semantic topics of lecture videos by considering audio and visual information. Specifically, the speaking content and the information of presentation slides are extracted from the lecture videos. With the proposed inference and learning algorithm, the semantic topics of the video can be discovered. The experimental results show that the proposed method can effectively discover the meaningful semantic characters of the lecture videos.