作者: Jeyhan Karaoguz , James D. Bennett
DOI:
关键词: Multiview Video Coding 、 Data compression 、 Reference frame 、 Frame (networking) 、 Block-matching algorithm 、 Encoding (memory) 、 Video compression picture types 、 Motion compensation 、 Computer vision 、 Artificial intelligence 、 Computer science
摘要: Systems, methods and apparatuses are described herein for encoding a plurality of video frame sequences, wherein each sequence corresponds to different perspective view the same subject matter. In accordance with various embodiments, is performed in hierarchical manner that leverages referencing between frames ones sequences (so-called “external referencing”), but also allows encoded representations only subset be provided when less than all required support particular viewing mode display system capable displaying matter two-dimensional mode, three-dimensional or multi-view mode.