Adjacent spatial motion vector predictor candidates improvement
An example method of video decoding includes receiving a video bitstream comprising a plurality of blocks, including a current block. The method also includes generating a motion vector prediction (MVP) list for the current block by interleaving respective spatial motion vector predictors (SMVPs) from left and above neighboring blocks. Generating the MVP list includes inserting a first SMVP into the MVP list, the first SMVP corresponding to a left neighboring block that is left of the current block, and inserting a second SMVP into the MVP list, the second SMVP corresponding to a top neighboring block that is above the current block. The method further includes decoding the current block using an MVP from the MVP list.
1 . A method of video decoding, the method comprising:
receiving a video bitstream comprising a plurality of blocks, including a current block;
generating a motion vector prediction (MVP) list for the current block by interleaving respective spatial motion vector predictors (SMVPs) from left and above neighboring blocks, including:
inserting a first SMVP into the MVP list, the first SMVP corresponding to a left neighboring block that is left of the current block; and
inserting a second SMVP into the MVP list, the second SMVP corresponding to a top neighboring block that is above the current block; and
decoding the current block using an MVP from the MVP list.
2 . The method of claim 1 , further comprising pruning one or more candidates from the MVP list based on determining that the one or more candidates have a same SMVP.
3 . The method of claim 2 , wherein the pruning is performed on only a portion of the MVP list.
4 . The method of claim 1 , wherein generating the MVP further includes:
inserting a third SMVP into the MVP list, the third SMVP corresponding to the left neighboring block; and
inserting a fourth SMVP into the MVP list, the fourth SMVP corresponding to the top neighboring block.
5 . The method of claim 4 , wherein generating the MVP further includes:
inserting a fifth SMVP into the MVP list, the fifth SMVP corresponding to a bottom-left neighboring block; and
inserting a sixth SMVP into the MVP list, the sixth SMVP corresponding to a top-right neighboring block.
6 . The method of claim 5 , wherein generating the MVP further includes inserting a seventh SMVP into the MVP list, the seventh SMVP corresponding to a top-left neighboring block.
7 . The method of claim 1 , further comprising assigning respective weights to motion vectors (MVs) of the MVP list.
8 . The method of claim 7 , wherein one or more of the MVs in the MVP list are assigned a weight of zero.
9 . The method of claim 1 , wherein, when one or more conditions are satisfied, the first SMVP is inserted into the MVP list before the second SMVP.
10 . The method of claim 9 , wherein, when one or more conditions are satisfied, the first SMVP is inserted into the MVP list after the second SMVP.
11 . The method of claim 9 , wherein the one or more conditions correspond to one or more of: a width of the current block, and a height of the current block.
12 . The method of claim 9 , wherein the one or more conditions correspond to whether a left or above neighboring block is selected as an MVP candidate.
13 . A computing system, comprising:
control circuitry;
memory; and
one or more sets of instructions stored in the memory and configured for execution by the control circuitry, the one or more sets of instructions comprising instructions for:
receiving video data comprising a plurality of blocks corresponding to one or more pictures, the plurality of blocks including a current block;
generating a motion vector prediction (MVP) list for the current block by interleaving respective spatial motion vector predictors (SMVPs) from left and above neighboring blocks, including:
inserting a first SMVP into the MVP list, the first SMVP corresponding to a left neighboring block that is left of the current block; and
inserting a second SMVP into the MVP list, the second SMVP corresponding to a top neighboring block that is above the current block; and
encoding the current block using an MVP from the MVP list.
14 . The computing system of claim 13 , wherein generating the MVP further includes:
inserting a third SMVP into the MVP list, the third SMVP corresponding to the left neighboring block; and
inserting a fourth SMVP into the MVP list, the fourth SMVP corresponding to the top neighboring block.
15 . The computing system of claim 14 , wherein generating the MVP further includes:
inserting a fifth SMVP into the MVP list, the fifth SMVP corresponding to a bottom-left neighboring block; and
inserting a sixth SMVP into the MVP list, the sixth SMVP corresponding to a top-right neighboring block.
16 . The computing system of claim 15 , wherein generating the MVP further includes inserting a seventh SMVP into the MVP list, the seventh SMVP corresponding to a top-left neighboring block.
17 . The computing system of claim 13 , wherein the one or more sets of instructions further comprises instructions for assigning respective weights to motion vectors (MVs) of the MVP list.
18 . The computing system of claim 13 , wherein the one or more sets of instructions further comprises instructions for pruning one or more candidates from the MVP list based on determining that the one or more candidates have a same SMVP.
19 . The computing system of claim 13 , wherein, when one or more conditions are satisfied, the first SMVP is inserted into the MVP list before the second SMVP.
20 . A method of obtaining a video bitstream, the method comprising:
generating a video bitstream, including:
receiving video data comprising a plurality of blocks corresponding to one or more pictures, the plurality of blocks including a current block;
generating a motion vector prediction (MVP) list for the current block by interleaving respective spatial motion vector predictors (SMVPs) from left and above neighboring blocks, including:
inserting a first SMVP into the MVP list, the first SMVP corresponding to a left neighboring block that is left of the current block; and
inserting a second SMVP into the MVP list, the second SMVP corresponding to a top neighboring block that is above the current block; and
encoding the current block using an MVP from the MVP list; and
transmitting the video bitstream including the encoded current block.