Learning Video Manifold for Segmenting Crowd Events and Abnormality Detection

Abstract

AbstractThis paper addresses the problem of analyzing video events in crowded scenes. A novel manifold learning method is proposed to achieve visualization and modeling of video events in a low dimensional space. In the proposed approach, a video is considered as a trajectory of frames in a low-dimensional space. This low-dimensional representation of a video preserves the spatio-temporal property of a video as well as the characteristic of the video. Different tasks of video content analysis such as visualization, video event segmentation and abnormality detection are achieved by analyzing these video trajectories based on the Hausdorff distance similarity measure. We evaluate our proposed method on the state-of-the-art public data-sets containing different crowd events. Qualitative and quantitative results show the promising performance of the proposed method.KeywordsVideo SequenceVideo FrameSpectral ClusterManifold LearningCrowd EventThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.