Temporal merge candidates in merge candidate lists in video coding
A VVC-standard encoder and a VVC-standard decoder implement improvements over VVC and ECM in a number of regards: a temporal motion vector prediction candidate selection method utilizing relocation of a collocated CTU; a temporal motion vector prediction candidate selection method utilizing expanded selection range; a temporal motion vector prediction candidate selection method utilizing unconditional derivation of a scaled motion vector; a temporal motion vector prediction candidate selection method utilizing omission of scaling uni-predicted motion vectors to bi-predicted motion vectors; a temporal motion vector prediction candidate selection method utilizing multiple options in setting a reference picture index; a temporal motion vector prediction candidate selection method utilizing scaling factor offsetting; a merge candidate list building method omitting a temporal motion vector prediction candidate; and a picture reconstruction method utilizing motion information refinement.
1. A computing system, comprising:
one or more processors, and
memory communicatively coupled to the one or more processors, the memory storing a frontend application executable by the one or more processors that, when executed by the one or more processors, perform associated operations comprising:
selecting a plurality of motion candidates for a current coding unit (“CU”) of a current picture, wherein a Temporal Motion Vector Prediction candidate (“TMVP candidate”) is selected by setting a reference picture index of the TMVP candidate to a reference picture index having a scaling factor closest to 1,
wherein the scaling factor is determined based on a ratio between:
a picture order count (“POC”) difference between a reference picture of the current picture and the current picture, and
a POC difference between a collocated picture and a reference picture of the collocated picture.
2. The computing system of claim 1 , wherein the scaling factor comprises the ratio between:
the POC difference between the reference picture of the current picture and the current picture, and
the POC difference between the collocated picture and the reference picture of the collocated picture.
3. The computing system of claim 1 , further comprising reordering the plurality of merge candidates according to Adaptive Reordering of Merge Candidates (“ARMC”).
4. The computing system of claim 1 , wherein a plurality of TMVP candidates of the plurality of merge candidates are reordered by cost values based on template matching, wherein a template matching cost value of a merge candidate comprises a sum of absolute differences (“SAD”) between samples of a template of the current block and respective reference samples.
5. A method, comprising:
selecting a plurality of motion candidates for a current coding unit (“CU”) of a current picture, wherein a Temporal Motion Vector Prediction candidate (“TMVP candidate”) is selected by setting a reference picture index of the TMVP candidate to a reference picture index having a scaling factor closest to 1,
wherein the scaling factor is determined based on a ratio between:
a picture order count (“POC”) difference between a reference picture of the current picture and the current picture, and
a POC difference between a collocated picture and a reference picture of the collocated picture.
6. The method of claim 5 , wherein the scaling factor comprises the ratio between:
the POC difference between the reference picture of the current picture and the current picture, and
the POC difference between the collocated picture and the reference picture of the collocated picture.
7. The method of claim 5 , further comprising reordering the plurality of merge candidates according to Adaptive Reordering of Merge Candidates (“ARMC”).
8. The method of claim 5 , wherein a plurality of TMVP candidates of the plurality of merge candidates are reordered by cost values based on template matching, wherein a template matching cost value of a merge candidate comprises a sum of absolute differences (“SAD”) between samples of a template of the current block and respective reference samples.
9. A non-transitory computer-readable storage medium storing a bitstream of a video for processing according to a method comprising:
parsing, by one or more processors of a computing system and in response to receiving the bitstream, a merge index signaled in a syntax structure of a coding unit (“CU”);
selecting, by the one or more processors and in response to receiving the bitstream, a plurality of motion candidates for the CU, wherein a Temporal Motion Vector Prediction candidate (“TMVP candidate”) is selected by setting a reference picture index of the TMVP candidate to a reference picture index having a scaling factor closest to 1,
wherein the scaling factor is determined based on a ratio between:
a picture order count (“POC”) difference between a reference picture of the current picture and the current picture, and
a POC difference between a collocated picture and a reference picture of the collocated picture; and
performing, by the one or more processors and in response to receiving the bitstream, inter prediction of the CU by applying merge mode based on the merge index and the plurality of motion candidates.
10. The non-transitory computer-readable storage medium of claim 9 , wherein the scaling factor comprises the ratio between:
the POC difference between the reference picture of the current picture and the current picture, and
the POC difference between the collocated picture and the reference picture of the collocated picture.
11. The non-transitory computer-readable storage medium of claim 9 , wherein the method further comprises reordering the plurality of merge candidates according to Adaptive Reordering of Merge Candidates (“ARMC”).
12. The non-transitory computer-readable storage medium of claim 9 , wherein a plurality of TMVP candidates of the plurality of merge candidates are reordered by cost values based on template matching, wherein a template matching cost value of a merge candidate comprises a sum of absolute differences (“SAD”) between samples of a template of the current block and respective reference samples.