A REVIEW ON VIDEO COMPRESSION APPROACHES AND UTILIZATION OF DEEP LEARNING TECHNIQUES

In this paper, an extensive review on video encoding-decoding schemes have been discussed. Further, a novel architecture has been laid off for suitable future work in the same direction. The reviews on literature plies from 1960 until date show casing the bench mark methods proposed by eminent researchers in the domain of video compression. The time line of representing the review is split into three categories. These categories refer to the classical methods, the conventional heuristic methods, and modern deep learning techniques utilized for the purpose of video compression. For all the categories, the milestone contributions are discussed. The approaches are summarized in different tables along with their advantages and disadvantages. Certain remarks for specific approaches are also provided in the summary. The limitations of existing works are well described in depth so that prospective researchers can have a path for future works. Finally, concluding remark along with prospective work in the same direction is presented.