Prediction refinement for affine merge and affine motion vector prediction mode

Patent metadata

Field	Value
Publication number	US-12513291-B2
Application number	US-202318516447-A
Country	US
Kind code	B2
Filing date	Nov 21, 2023
Priority date	Mar 23, 2020
Publication date	Dec 30, 2025
Grant date	Dec 30, 2025

Abstract

A method includes determining, for a conversion between a video block of a video and a bitstream of the video, a size of prediction block corresponding to the video block according to a rule. The method also includes performing the conversion based on the determining. The rule specifies that a first size of the prediction block is determined responsive to whether a prediction refinement using optical flow technique is used for coding the video block. The video block has a second size and is coded using an affine merge mode or an affine advanced motion vector prediction mode.

First claim

What is claimed is:

1 . A method of processing video data, comprising:
performing a first conversion between a video comprising a picture comprising one or more slices and a bitstream of the video, wherein the first conversion conforms to a first rule that specifies that whether a deblocking filter is applied to a first slice referring to a picture parameter set (PPS) is based at least one of: a first syntax element in a slice header (SH) indicating whether an override of information of the deblocking filter is enabled at a slice level, or a second syntax element in a picture header (PH) indicating whether the override of information of the deblocking filter is enabled at a picture level; and wherein the first slice is one of the one or more slices comprised in the picture; determining, for a second conversion between a first video block of the video and the bitstream of the video, a size of a prediction block corresponding to the first video block according to a second rule; and performing the second conversion between the first video block of the video and the bitstream of the video based on the determining, wherein an affine merge mode is enabled for the first video block, wherein the second rule specifies that a first size of the prediction block is determined responsive to whether a prediction refinement using optical flow technique is enabled for the first video block, and wherein the first video block has a second size; wherein a first width and a first height of the first size of the prediction block are indicated by (M+M0) and (N+N0), respectively, wherein a second width and a second height of the second size of the first video block are indicated by M and N, respectively, and wherein M, M0, N and N0 are integers greater than or equal to 0.

2 . The method of claim 1 , wherein the first rule specifies that whether the deblocking filter is applied to the first slice referring to the picture parameter set is further based at least on a third syntax element included in the PH associated with the first slice, and
wherein the third syntax element indicates whether the deblocking filter is disabled for the picture.

3 . The method of claim 2 , wherein the first rule specifies that whether the deblocking filter is applied to the one or more slices referring to the picture parameter set is further based on at least one of:
a fourth syntax element in the PPS indicating whether the deblocking filter is disabled for pictures referring to the PPS, or a fifth syntax element at the slice level indicating whether the deblocking filter is indicated as being disabled.

4 . The method of claim 3 , wherein values of the third syntax element, the fourth syntax element, and the fifth syntax element equal to 1 indicate that the deblocking filter is disabled at the picture level, a PPS level, and the slice level, respectively.

5 . The method of claim 4 , wherein when the values of the fourth syntax element and the fifth syntax element equal to 1, the deblocking filter is disabled for the first slice.

6 . The method of claim 2 , wherein the first rule further specifies that a presence of a fifth syntax element in the SH and/or the third syntax element in the PH is based on a value of a fourth syntax element in the PPS.

7 . The method of claim 6 , wherein the first rule further specifies that the deblocking filter is enabled on the first slice when the fourth syntax element indicates that the deblocking filter is disabled for the picture.

8 . The method of claim 7 , wherein the first rule further specifies that for multiple slices in the picture, the deblocking filter is enabled for some slices and disabled for other slices.

9 . The method of claim 1 , wherein when the prediction refinement using optical flow technique is enabled for the first video block, at least one of M0 and N0 is not equal to 0; and
wherein M0 and N0 are both equal to 2.

10 . The method of claim 1 , wherein a prediction refinement utility flag controls values of M0 and N0, and wherein the prediction refinement utility flag indicates whether the prediction refinement using optical flow technique is utilized; and
wherein the values of M0 and N0 are determined independently from an affine flag.

11 . The method of claim 10 , wherein the affine flag is inter_affine_flag which is used to indicate whether to apply an affine motion vector prediction mode; and
wherein for a video block applied with the affine motion vector prediction mode, a size of a prediction block of the video block is equal to the first size.

12 . The method of claim 9 , wherein a prediction sample of the prediction block is present as
predSamplesLX [xL][yL], wherein xL is between 0 and M+1 inclusive, and yL is between 0 and N+1 inclusive, and wherein the prediction sample predSamplesLX [xL][yL] is derived by invoking a luma integer sample fetching process when one or more of conditions are true: xL is equal to 0, xL is equal to M+1, yL is equal to 0 and yL is equal to N+1, and the prediction sample predSamplesLX [xL][yL] is derived by invoking a luma sample 8-tap interpolation filtering process for all conditions that are false.

13 . The method of claim 1 , wherein the affine merge mode includes generating control point motion vectors by using a merge index to select an affine merge candidate from a sub-block merge candidate list which is constructed based on motion information of spatial neighboring coding units.

14 . The method of claim 1 , wherein the performing the first conversion comprises encoding the video into the bitstream.

15 . The method of claim 1 , wherein the performing the first conversion comprises decoding the video from the bitstream.

16 . 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:
perform a first conversion between a video comprising a picture comprising one or more slices and a bitstream of the video, wherein the first conversion conforms to a first rule that specifies that whether a deblocking filter is applied to a first slice referring to a picture parameter set (PPS) is based at least one of: a first syntax element in a slice header (SH) indicating whether an override of information of the deblocking filter is enabled at a slice level, or a second syntax element in a picture header (PH) indicating whether the override of information of the deblocking filter is enabled at a picture level; and wherein the first slice is one of the one or more slices comprised in the picture; determine, for a second conversion between a first video block of the video and the bitstream of the video, a size of a prediction block corresponding to the first video block according to a second rule; and perform the second conversion between the first video block of the video and the bitstream of the video based on the determining, wherein an affine merge mode is enabled for the first video block, wherein the second rule specifies that a first size of the prediction block is determined responsive to whether a prediction refinement using optical flow technique is enabled for the first video block, and wherein the first video block has a second size; wherein a first width and a first height of the first size of the prediction block are indicated by (M+M0) and (N+N0), respectively, wherein a second width and a second height of the second size of the first video block are indicated by M and N, respectively, and wherein M, M0, N and N0 are integers greater than o

Assignees

• Beijing Bytedance Network Tech Co Ltd
• Bytedance Inc

Inventors

• Deng Zhipin
• Zhang Li
• Wang Ye-Kui
• Zhang Kai
• Xu Jizheng

Classifications

• H04N19/80

  Details of filtering operations specially adapted for video compression, e.g. for pixel interpolation (H04N19/635, H04N19/86 take precedence) · CPC title

• H04N19/52

  by predictive encoding · CPC title

• H04N19/463

  by compressing encoding parameters before transmission · CPC title

• H04N19/186

  the unit being a colour or a chrominance component · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title