A tensor framework for data stream clustering and compression

Abstract

In the paper a tensor based method for video stream clustering and compression is presented. The method does video partitioning in temporal domain based on its content. Such coherent video partitions are amenable for better compression. The proposed method detects shot boundaries building a tensor model from a number of frames in the stream. To build the model, the best rank tensor decomposition is used. Each incoming tensor-frame is verified with the model based on the proposed concept drift detector – if it fits, then the model is updated with that frame. Otherwise, a model is rebuilt. This way obtained shots are then compressed also with the best rank tensor decomposition methods.