CPC H04N 19/105 (2014.11) [H04L 65/70 (2022.05); H04L 65/75 (2022.05); H04N 19/137 (2014.11); H04N 19/174 (2014.11); H04N 19/176 (2014.11); H04N 19/184 (2014.11); H04N 19/513 (2014.11); H04N 19/70 (2014.11); H04N 19/96 (2014.11)] | 20 Claims |
1. A method for processing video data, comprising:
maintaining one or multiple tables, wherein each table includes one or multiple motion candidates derived from one or more video blocks that have been coded, and arrangement of the motion candidates in the each table is based on a sequence of addition of the motion candidates into the each table, each motion candidate comprises motion information associated with M representative positions of the video block that has been coded;
performing a conversion between a current video block and a coded representation of the current video block at least based on a motion candidate list;
determining M representative positions within the current video block, where M is an integer, to load motion information associated with the M representative positions; and
updating at least one of the one or multiple tables using motion information associated with the M representative positions of the current video block;
wherein the updated table used to construct a motion candidate list for a subsequent block, and wherein the motion candidate list is used for deriving motion information of the subsequent block;
wherein, in a case sub-block predictions are allowed for processing of the current video block, a value of M is set to 1, and the M representative positions are selected based on at least one of a frequency of utilized motion information, whether a bi-prediction block is used, a reference picture, a reference picture index, or motion vector differences compared to other motion vectors.
|