Generating Optimal Video Summaries

In this paper, we propose a novel technique for video summarization based on the Singular Value Decomposition (SVD). For the input video sequence, we create a featureframe matrix A, and perform the SVD on it. From this SVD, we are able to derive not only the refined feature space to better cluster visually similar frames, but also a metric to measure the amount of visual content contained in each frame cluster using its degree of visual changes. Then, in this refined feature space, we find the most static frame cluster, define it as the content unit, and use the content value computed from this cluster as well as the distance between frames as the thresholds to cluster the rest of the frames. Based on this clustering result, the optimal set of keyframes is generated as the content summary of the original video. Our approach ensures that the summarized video representation contains little redundancy, and gives equal attention to the same amount of visual contents. Besides the optim...

• Type: Conference
• Year: 2000
• Where: ICMCS
• Authors: Yihong Gong, Xin Liu
• Keywords: Frame Cluster, ICMCS 2000, Input Video Sequence, Multimedia, Summarized Motion Videos