Video summary quality evaluation based on 4C assessment and user interaction

As video summarization techniques have attracted increasing attention for efficient multimedia data management, quality evaluation of video summary is required. To address the lack of automatic evaluation techniques, this chapter proposes a novel full-reference evaluation framework to assess the quality of the video summary according to various user requirements. First, the reference video summary and the candidate video summary are decomposed into two sequences of Summary Units (SUs), and the SUs in these two sequences are matched by frame alignment. Then, a similarity-based assessment algorithm is proposed to automatically provide comprehensive human-like evaluation results of the candidate video summary quality from the perspective of Coverage, Conciseness, Coherence, and Context (4C), respectively. Considering the evaluation, criteria of video summary quality are usually application-dependent, the incremental user interaction is utilized to gather the user requirements of video summary quality, and the required evaluation results are transformed from the 4C assessment scores. The proposed framework is experimented on a standard dataset of TRECVID 2007 and shows a good performance in automatic video summary evaluation.