Bi-directional optical flow and decoder-side motion vector refinement for video coding
Methods and devices are provided for deriving constructed affine merge candidates. The method includes obtaining a first reference picture and a second reference picture associated with an inter mode coded block, where the first reference picture is before a current picture and the second reference picture is after the current picture in display order, obtaining a first motion vector from the inter mode coded block to a reference block in the first reference picture, obtaining a second motion vector from the inter mode coded block to a reference block in the second reference picture, applying bi-directional optical flow (BDOF) or decoder-side motion vector refinement (DMVR) to the inter mode coded block based on a mode information of the inter mode coded block, and predicting a bi-prediction of the inter mode coded block based on the applied BDOF or DMVR.
1. A method of coding a video signal, comprising:
obtaining a first reference picture and a second reference picture associated with an inter mode coded block, wherein the first reference picture is before a current picture and the second reference picture is after the current picture in display order;
obtaining a first motion vector from the inter mode coded block to a reference block in the first reference picture;
obtaining a second motion vector from the inter mode coded block to a reference block in the second reference picture;
applying bi-directional optical flow (BDOF) or decoder-side motion vector refinement (DMVR) to the inter mode coded block based on a mode information of the inter mode coded block, wherein the mode information includes the first motion vector and second motion vector; and
predicting a bi-prediction of the inter mode coded block based on the applied BDOF or DMVR,
wherein applying the BDOF or the DMVR to the inter mode coded block based on the mode information of the inter mode coded block comprises:
determining whether a regular merge mode and a coded merge index is chosen for the inter mode coded block;
applying the DMVR in response to the regular merge mode having an odd merge index number and being used on the inter mode coded block; and
applying the BDOF in response to the regular merge mode having an even merge index number and being used on the inter mode coded block.
2. The method of claim 1 , further comprising:
providing a flag specifying whether the BDOF or the DMVR is applied to the inter mode coded block.
3. The method of claim 1 , further comprising:
generating a merge candidate list for the BDOF and providing a merge index in response to the merge candidate list being greater than one;
providing a merge syntax to indicate a BDOF merge mode; and
disabling the BDOF in response to the DMVR being applied.
4. The method of claim 1 , further comprising:
generating a merge candidate list for the DMVR and providing a merge index in response to the merge candidate list being greater than one;
providing a merge syntax to indicate a DMVR merge mode; and
applying the DMVR.
5. A computing device, comprising:
one or more processors;
a non-transitory computer-readable storage medium storing instructions executable by the one or more processors, wherein the one or more processors are configured to:
obtain a first reference picture and a second reference picture associated with an inter mode coded block, wherein the first reference picture is before a current picture and the second reference picture is after the current picture in display order;
obtain a first motion vector from the inter mode coded block to a reference block in the first reference picture;
obtain a second motion vector from the inter mode coded block to a reference block in the second reference picture;
apply bi-directional optical flow (BDOF) or decoder-side motion vector refinement (DMVR) to the inter mode coded block based on a mode information of the inter mode coded block, wherein the mode information includes the first motion vector and second motion vector; and
predict a bi-prediction of the inter mode coded block based on the applied BDOF or DMVR,
wherein the one or more processors configured to apply the BDOF or the DMVR to the inter mode coded block based on the mode information of the inter mode coded block are further configured to:
determine whether a regular merge mode or a coded merge index is chosen for the inter mode coded block;
apply the DMVR in response to the regular merge mode having an odd merge index number and being used on the inter mode coded block; and
apply the BDOF in response to the regular merge mode having an even merge index number and being used on the inter mode coded block.
6. A non-transitory computer-readable storage medium storing a plurality of programs for execution by a computing device having one or more processors, wherein the plurality of programs, when executed by the one or more processors, cause the computing device to perform acts comprising:
obtaining a first reference picture and a second reference picture associated with an inter mode coded block, wherein the first reference picture is before a current picture and the second reference picture is after the current picture in display order;
obtaining a first motion vector from the inter mode coded block to a reference block in the first reference picture;
obtaining a second motion vector from the inter mode coded block to a reference block in the second reference picture;
applying bi-directional optical flow (BDOF) or decoder-side motion vector refinement (DMVR) to the inter mode coded block based on a mode information of the inter mode coded block, wherein the mode information includes the first motion vector and second motion vector; and
predicting a bi-prediction of the inter mode coded block based on the applied BDOF or DMVR,
wherein applying the BDOF or the DMVR to the inter mode coded block based on the mode information of the inter mode coded block comprises:
determining whether a regular merge mode and a coded merge index is chosen for the inter mode coded block;
applying the DMVR in response to the regular merge mode having an odd merge index number and being used on the inter mode coded block; and
applying the BDOF in response to the regular merge mode having an even merge index number and being used on the inter mode coded block.
7. The computing device of claim 5 , wherein the one or more processors are further configured to:
provide a flag specifying whether the BDOF or the DMVR is applied to the inter mode coded block.
8. The computing device of claim 5 , wherein the one or more processors are further configured to:
generate a merge candidate list for the BDOF and provide a merge index in response to the merge candidate list being greater than one;
provide a merge syntax to indicate a BDOF merge mode; and
disable the BDOF in response to the DMVR being applied.
9. The computing device of claim 5 , wherein the one or more processors are further configured to:
generate a merge candidate list for the DMVR and provide a merge index in response to the merge candidate list being greater than one;
provide a merge syntax to indicate a DMVR merge mode; and
apply the DMVR.
10. The medium of claim 6 , wherein the acts further comprising:
providing a flag specifying whether the BDOF or the DMVR is applied to the inter mode coded block.
11. The medium of claim 6 , wherein the acts further comprising:
generating a merge candidate list for the BDOF and providing a merge index in response to the merge candidate list being greater than one;
providing a merge syntax to indicate a BDOF merge mode; and
disabling the BDOF in response to the DMVR being applied.
12. The medium of claim 6 , wherein the acts further comprising:
generating a merge candidate list for the DMVR and providing a merge index in response to the merge candidate list being greater than one;
providing a merge syntax to indicate a DMVR merge mode; and
applying the DMVR.