Motion precision in sub-block based inter prediction
Devices, systems and methods for digital video coding, which include sub-block based inter prediction methods, are described. An exemplary method for video processing includes determining, for a conversion between a current block of video and a bitstream representation of the video, a maximum number of candidates in a sub-block based merge candidate list and/or whether to add sub-block based temporal motion vector prediction (SbTMVP) candidates to the sub-block based merge candidate list based on whether temporal motion vector prediction (TMVP) is enabled for use during the conversion or whether a current picture referencing (CPR) coding mode is used for the conversion, and performing, based on the determining, the conversion.
1. A method of video processing, comprising:
determining, for a conversion between a current block of a video coded using a sub-block based temporal motion vector prediction tool and a bitstream of the video, a temporal motion vector;
modifying the temporal motion vector by right-shifting without left-shifting the temporal motion vector to an integer precision, wherein the right-shifting is a right-shifting by a fixed value equal to 4;
locating at least one region for the current block in a collocated picture based on the modified temporal motion vector, wherein the at least one region includes multiple pixels with integer precision; and
performing the conversion based on the at least one region,
wherein performing the conversion comprises:
acquiring a temporal motion vector prediction based on the at least one region;
wherein the determining the temporal motion vector comprises:
initializing the temporal motion vector to a default motion vector, and
setting the temporal motion vector to a specific motion vector related to a specific spatial neighboring block of the current block, wherein the specific spatial neighboring block is in a current frame comprising the current block.
2. The method of claim 1 , wherein the right-shifting uses a rounding toward zero operation.
3. The method of claim 1 , wherein the performing the conversion based on the at least one region comprises:
deriving at least one sub-block motion information related to the current block based on the at least one region;
constructing a sub-block motion candidate list at least based on the at least one sub-block motion information; and
performing the conversion based on the sub-block motion candidate list, wherein a sub-block merge index is included in the bitstream.
4. The method of claim 3 , wherein a plurality of bins (M) is used to present the sub-block merge index, wherein a first number of bins (L) of the plurality of bins are context coded, and wherein a second number of bins (M−L) are bypass coded, wherein L=1.
5. The method of claim 1 , wherein a motion vector related to a default motion information is (0, 0).
6. The method of claim 1 , wherein the right-shifting uses a rounding operation that is identical to that used in a motion vector averaging process, wherein in the motion vector averaging process, a pairwise merge candidate is derived.
7. The method of claim 1 , wherein the right-shifting uses a rounding operation that is identical to that used in a motion vector scaling process or an adaptive motion vector resolution process.
8. The method of claim 1 , wherein the conversion comprises decoding the current block from the bitstream.
9. The method of claim 1 , wherein the conversion comprises encoding the current block into the bitstream.
10. The method of claim 1 , wherein, in response to the specific spatial neighboring block being coded with an inter mode, the temporal motion vector is set to the specific motion vector related to the specific spatial neighboring block of the current block.
11. The method of claim 1 , wherein the performing the conversion based on the at least one region comprises:
if the at least one region is coded with an inter mode, performing the conversion based on the at least one region and using the sub-block based temporal motion vector prediction tool.
12. An apparatus for processing video data comprising a processor and a non-transitory memory with instructions thereon, wherein the instructions upon execution by the processor, cause the processor to:
determine, for a conversion between a current block of a video coded using a sub-block based temporal motion vector prediction tool and a bitstream of the video, a temporal motion vector;
modify the temporal motion vector by right-shifting without left-shifting the temporal motion vector to an integer precision, wherein the right-shifting is a right-shifting by a fixed value equal to 4;
locate at least one region for the current block in a collocated picture based on the modified temporal motion vector, wherein the at least one region includes multiple pixels with integer precision; and
perform the conversion based on the at least one region,
wherein performing the conversion comprises:
acquiring a temporal motion vector prediction based on the at least one region,
wherein the determining the temporal motion vector comprises:
initializing the temporal motion vector to a default motion vector, and
setting the temporal motion vector to a specific motion vector related to a specific spatial neighboring block of the current block, wherein the specific spatial neighboring block is in a current frame comprising the current block.
13. The apparatus of claim 12 , wherein the right-shifting uses a rounding toward zero operation.
14. A non-transitory computer-readable storage medium storing instructions that cause a processor to:
determine, for a conversion between a current block of a video coded using a sub-block based temporal motion vector prediction tool and a bitstream of the video, a temporal motion vector;
modify the temporal motion vector by right-shifting without left-shifting the temporal motion vector to an integer precision, wherein the right-shifting is a right-shifting by a fixed value equal to 4;
locate at least one region for the current block in a collocated picture based on the modified temporal motion vector, wherein the at least one region includes multiple pixels with integer precision; and
perform the conversion based on the at least one region,
wherein performing the conversion comprises:
acquiring a temporal motion vector prediction based on the at least one region,
wherein the determining the temporal motion vector comprises:
initializing the temporal motion vector to a default motion vector, and
setting the temporal motion vector to a specific motion vector related to a specific spatial neighboring block of the current block, wherein the specific spatial neighboring block is in a current frame comprising the current block.
15. A method for storing a bitstream of a video, comprising:
determining, for a current block of the video coded using a sub-block based temporal motion vector prediction tool, a temporal motion vector;
modifying the temporal motion vector by right-shifting without left-shifting the temporal motion vector to an integer precision, wherein the right-shifting is a right-shifting by a fixed value equal to 4;
locating at least one region for the current block in a collocated picture based on the modified temporal motion vector, wherein the at least one region includes multiple pixels with integer precision;
generating the bitstream based on the at least one region,
wherein generating the bitstream comprises:
acquiring a temporal motion vector prediction based on the at least one region; and
storing the bitstream in a non-transitory computer-readable recording medium,
wherein the determining the temporal motion vector comprises:
initializing the temporal motion vector to a default motion vector, and
setting the temporal motion vector to a specific motion vector related to a specific spatial neighboring block of the current block, wherein the specific spatial neighboring block is in a current frame comprising the current block.