Video summarization aims to provide a compact video representation while preserving the essential activities of the\noriginal video. Most existing video summarization approaches relay on identifying important frames and optimizing\ntarget energy by a global optimum solution. But global optimum may fail to express continuous action or realistically\nvalidate how human beings perceive a story. In this paper, we present a bottom-up approach named clip growing for\nvideo summarization, which allows users to customize the quality of the video summaries. The proposed approach\nfirstly uses clustering to oversegment video frames into video clips based on their similarity and proximity.\nSimultaneously, the importance of frames and clips is evaluated from their corresponding dissimilarity and\nrepresentativeness. Then, video clips and frames are gradually selected according to their energy rank, until reaching\nthe target length. Experimental results on SumMe dataset show that our algorithm can produce promising results\ncompared to existing algorithms. Several video summarizations results are presented in supplementary material.
Loading....