Semantic browsing in large scale videos collection

With the increasing amount of video data being produced, there is a growing demand for more efficient tools of exploration and navigation to further improve the retrieval effectiveness. In this paper, we present a new approach for video browsing based on semantic and visual suggestions methodology. First, we describe a new method to build Semantic Concepts Network (SCN), which combines conceptual and contextual relationships among concepts. Second, we explore semantic links in SCN to propose new concepts and video shots that seems to be relevant for the user intent. A user study has shown that the proposed approach is able to help the user for semantic browsing in video collection.