Video summarization using viewer affective feedback

For different reasons, many viewers like to watch a summary of films in less time than it takes to play. Traditionally, video films were analyzed manually to provide its summary; however this requires an important work time. Therefore, it is a necessity to propose a tool for summarization and video analysis automatically. Automatic video summarization has to extract all important moments in which viewers may be interested. All summarization criteria can be different from a video to another. In a different context with foregoing video summarization works, this paper present how the emotional dimensions issued from the real viewers can be considered as the important input to compute which part is most interesting in the total time of a film.