Manipulating temporal dependencies in compressed video data with applications to compressed-domain processing of MPEG video

The ability to manipulate the temporal dependencies in coded video data is important for a number of compressed domain video processing tasks. This paper formulates the general problem and examines it in the context of MPEG. This is used to develop a method for performing frame conversions in MPEG coded video data. These frame conversions are used to develop compressed-domain video processing algorithms for performing temporal mode conversion, frame-by-frame reverse play, and frame-accurate splicing.