HIGH LEVEL SYNTAX IMPROVEMENT ON INTER-LAYER PREDICTION FOR SHVC/MV-HEVC

摘要

Syntax changes for SHVC/MV-HEVC coding apparatus and methods are described which improve efficiency. In a first portion, syntax of header signaling position under P or B-slice conditions are described, and a new condition to signal inter-layer prediction layer. A second portion describes use of an inter-layer reference picture set as an alternative to the first portion, and provides an arranged syntax to more efficiently signal inter-layer reference information. In a third portion, an inter-layer sample only prediction and TMVP interaction is described which reduces signaling needs for collocated references syntax.

主权项

1. A video coding apparatus, comprising:
(a) at least one computer processor; and (b) programming executable on said at least one computer processor for removing redundancy in an inter-layer operations by performing steps comprising:
(i) removing redundancy in an inter-layer prediction syntax by configuring inter-layer prediction so that it is only performed for P or B types of slices, and performing inter-layer prediction when the number of reference layers differs between the sequence and picture level; and(ii) eliminating unnecessary temporal motion vector prediction (TMVP), by generating collocated reference syntax only if inter-layer prediction flag is false, whereby TMVP is disabled on a current picture when only inter-layer reference pictures exist in reference picture lists of all slices in the current picture, because motion vectors from inter layer reference pictures are constrained to be zero motion only.