Abstract
With the fast development and popularity of digital cameras, smart phones, and video surveillance devices, the amount of video data increases dramatically. Accordingly, automatically mining and annotating high-level concepts for video indexing and management become imperative research tasks in both multimedia research and data mining research. The mainstream content-based semantic concept mining approaches suffer from the notorious semantic gap problem, which is the difficulty of associating the low-level features to high-level concepts directly. Recently, the utilization of concept-concept association has been proven to be effective to address the semantic gap problem. In this paper, a framework based on multi-model collaboration and information integration is proposed to integrate the association among concepts to enhance the high-level concept detection by reusing the information outputs from the primary concept detectors. The experimental results show that the proposed framework outperforms the other approaches in the comparison and therefore is promising.