Temporal Tessellation for Video Annotation and Summarization

Abstract

We present a general approach to video understanding, inspired by semantic transfer techniques successfully used for 2D image understanding. Our method considers a video to be a 1D sequence of clips, each one associated with its own semantics. The nature of these semantics -- natural language captions or other labels -- depends on the task at hand. A test video is processed by forming correspondences between its clips and the clips of reference videos with known semantics, following which, reference semantics can be transferred to the test video. We describe two matching methods, both designed to ensure that (a) reference clips appear similar to test clips and (b), taken together, the semantics of selected reference clips is consistent and maintains temporal coherence. We use our method for video captioning on the LSMDC'16 benchmark and video summarization on the SumMe benchmark. In both cases, our method not only surpasses state of the art results, but importantly, it is the only method we know of that was successfully applied to both video understanding tasks.