Sub-block motion derivation and decoder-side motion vector refinement for merge mode
Systems, methods, and instrumentalities for sub-block motion derivation and motion vector refinement for merge mode may be disclosed herein. Video data may be coded (e.g., encoded and/or decoded). A collocated picture for a current slice of the video data may be identified. The current slice may include one or more coding units (CUs). One or more neighboring CUs may be identified for a current CU. A neighboring CU (e.g., each neighboring CU) may correspond to a reference picture. A (e.g., one) neighboring CU may be selected to be a candidate neighboring CU based on the reference pictures and the collocated picture. A motion vector (MV) (e.g., collocated MV) may be identified from the collocated picture based on an MV (e.g., a reference MV) of the candidate neighboring CU. The current CU may be coded (e.g., encoded and/or decoded) using the collocated MV.
1. A device for video decoding, comprising:
a processor configured to:
obtain a collocated picture associated with a current video block;
determine a constrained region in the collocated picture based on a location of the current video block, wherein a size of the constrained region is greater than a size of the current video block;
determine a first location associated with a collocated block based on the location of the current video block and a temporal motion vector (MV);
determine whether the first location associated with the collocated block is within the constrained region in the collocated picture, wherein, based on the first location associated with the collocated block being outside the constrained region, the processor is configured to determine a second location associated with the collocated block using a clipping operation associated with the constrained region in the collocated picture, wherein, the clipping operation changes the first location associated with the collocated block based on a boundary of the constrained region;
obtain, in the collocated picture, the collocated block associated with the current video block at the second location associated with the collocated block;
obtain an MV associated with the collocated block; and
decode the current video block based on the MV associated with the collocated block.
2. The device of claim 1 , wherein the current video block is an advanced temporal motion vector prediction (ATMVP) subblock, and the collocated block is a collocated subblock associated with the ATMVP subblock, and wherein the processor is further configured to:
obtain an MV associated with the collocated subblock; and
predict the ATMVP subblock based on the MV associated with the collocated subblock, wherein the current video block is decoded based on the prediction of the ATMVP subblock.
3. The device of claim 1 , wherein the processor is further configured:
determine a picture order count (POC) difference between the collocated picture and a reference picture of a neighboring block of the current video block; and
determine the temporal MV based on the POC difference, wherein the constrained region in the collocated picture is determined further based on the temporal MV.
4. The device of claim 1 , wherein a coding tree block (CTU) comprises the current video block, and the constrained region is further determined based on a location of the CTU in the collocated picture.
5. The device of claim 1 , wherein a coding tree block (CTU) comprises the current video block, and wherein an area of the constrained region is determined to be the same size as an area of the CTU plus one column of a video block.
6. The device of claim 1 , wherein a coding tree block (CTU) comprises the current video block, and wherein an area of the constrained region is determined to be the same size as an area of the CTU.
7. The device of claim 1 , wherein the MV of the collocated block is associated with a reference picture of the collocated block.
8. The device of claim 1 , wherein the second location associated with the collocated block is on the boundary of the constrained region.
9. A method for video decoding, comprising:
obtaining a collocated picture associated with a current video block;
determining a constrained region in the collocated picture based on a location of the current video block, wherein a size of the constrained region is greater than a size of the current video block;
determining a first location associated with a collocated block based on the location of the current video block and a temporal motion vector (MV);
determining whether the first location associated with the collocated block is within the constrained region in the collocated picture, wherein, based on the first location associated with the collocated block being outside the constrained region, the method comprises determining a second location associated with the collocated block using a clipping operation associated with the constrained region in the collocated picture, wherein the clipping operation changes the first location associated with the collocated block based on a boundary of the constrained region;
obtaining, in the collocated picture, the collocated block associated with the current video block at the second location associated with the collocated block;
obtaining an MV associated with the collocated block; and
decoding the current video block based on the MV associated with the collocated block.
10. The method of claim 9 , wherein the MV of the collocated block is associated with a reference picture of the collocated block.
11. The method of claim 9 , wherein the current video block is an advanced temporal motion vector prediction (ATMVP) subblock, and the collocated block is a collocated subblock associated with the ATMVP subblock, and wherein the method further comprises:
obtaining an MV associated with the collocated subblock; and
predicting the ATMVP subblock based on the MV associated with the collocated subblock, wherein the current video block is decoded based on the prediction of the ATMVP subblock.
12. The method of claim 9 , wherein the second location associated with the collocated block is on the boundary of the constrained region.
13. A device for video encoding, comprising:
a processor configured to:
obtain a collocated picture associated with a current video block;
determine a constrained region in the collocated picture based on a location of the current video block, wherein a size of the constrained region is greater than a size of the current video block;
determine a first location associated with a collocated block based on the location of the current video block and a temporal motion vector (MV);
determine whether the first location associated with the collocated block is within the constrained region in the collocated picture, wherein, based on the first location associated with the collocated block being outside the constrained region, the processor is configured to determine a second location associated with the collocated block using a clipping operation associated with the constrained region in the collocated picture, wherein the clipping operation changes the first location associated with the collocated block based on a boundary of the constrained region;
obtain, in the collocated picture, the collocated block associated with the current video block at the second location associated with the collocated block;
obtain an MV associated with the collocated block; and
encode the current video block based on the MV associated with the collocated block.
14. The device of claim 13 , wherein the MV of the collocated block is associated with a reference picture of the collocated block.
15. The device of claim 13 , wherein the current video block is an advanced temporal motion vector prediction (ATMVP) subblock, and the collocated block is a collocated subblock associated with the ATMVP subblock, and wherein the processor is further configured to:
obtain an MV associated with the collocated subblock; and
predict the ATMVP subblock based on the MV associated with the collocated subblock, wherein the current video block is encoded based on the prediction of the ATMVP subblock.
16. The device of claim 13 , wherein the second location associated with the collocated block is on the boundary of the constrained region.
17. A method for video encoding, comprising:
obtaining a collocated picture associated with a current video block;
determining a constrained region in the collocated picture based on a location of the current video block, wherein a size of the constrained region is greater than a size of the current video block;
determining a first location associated with a collocated block based on the location of the current video block and a temporal motion vector (MV);
determining whether the first location associated with the collocated block is within the constrained region in the collocated picture, wherein, based on the first location associated with the collocated block being outside the constrained region, the method comprises determining a second location associated with the collocated block using a clipping operation associated with the constrained region in the collocated picture, wherein the clipping operation changes the first location associated with the collocated block based on a boundary of the constrained region;
obtaining, in the collocated picture, the collocated block associated with the current video block at the second location associated with the collocated block;
obtaining an MV associated with the collocated block; and
encoding the current video block based on the MV associated with the collocated block.
18. The method of claim 17 , wherein the current video block is an advanced temporal motion vector prediction (ATMVP) subblock, and the collocated block is a collocated subblock associated with the ATMVP subblock, and wherein the method further comprises:
obtaining an MV associated with the collocated subblock; and
predicting the ATMVP subblock based on the MV associated with the collocated subblock, wherein the current video block is encoded based on the prediction of the ATMVP subblock.
19. The method of claim 17 , wherein the current video block is an advanced temporal motion vector prediction (ATMVP) subblock, and the collocated block is a collocated subblock associated with the ATMVP subblock, and wherein the method further comprises:
obtaining an MV associated with the collocated subblock; and
predicting the ATMVP subblock based on the MV associated with the collocated subblock, wherein the current video block is decoded based on the prediction of the ATMVP subblock.
20. The method of claim 17 , wherein the second location associated with the collocated block is on the boundary of the constrained region.