Generalized bi directional prediction mode in video processing

1. A method of processing video data, comprising:

performing a conversion between a current video block of a video and a bitstream of the video, based on a rule that controls a parameter of a generalized bi-prediction (GBi) coding tool in which a prediction of the current video block uses a final predictor corresponding to a non-uniformly weighted sum of predictors from two reference lists;

determining, for the conversion, a mode of operation of the GBi coding tool, and the conversion is performed based on the determining,

wherein the rule specifies to reorder weighting factors for the GBi coding tool based on a difference between neighboring pixels of the current video block and a result of averaging of corresponding reference neighboring pixels identified by motion vectors or integer part of the motion vectors of the current video block using the weighting factor, wherein the reference neighboring pixels are from the two reference lists,

wherein a field in the bitstream corresponds to the mode of operation and a value of the field indicates whether the GBI coding tool is enabled or disabled for the current video block,

wherein the GBi coding tool is disabled for the current video block that uses a quantization parameter greater than a threshold.

2. The method of claim 1 , wherein the field indicates that the GBI coding tool is either enabled or disabled at a slice level, a tile level, or a picture level, and the field is included in a slice header, a tile header, or a picture parameter set,

wherein in a case that a picture of the current video block corresponds to a temporal layer that has a value greater than a threshold, the field indicates that the GBi coding tool is disabled for the current video block, wherein the threshold is 3.

3. The method of claim 1 , wherein the rule specifies that a weighting factor set used by the GBi coding tool is based on i) a temporal layer of a picture including the current video block, ii) a picture quantization parameter of the picture, or iii) a quantization parameter of the current video block.

4. The method of claim 3 , wherein the rule specifies to use fewer weighting factors in higher temporal layer video pictures, optionally, the rule specifies to use fewer weighting factors for higher quantization parameter values.

5. The method of claim 3 , wherein the bitstream further includes at least one additional weighting factor set for use by the GBi coding tool, wherein the at least one additional weighting factor set is indicated at a slice header level, a tile group header level, a tile header level, a picture header level, a picture parameter set level, a sequence parameter set level, or a video parameter set level.

6. The method of claim 1 , further comprising:

determining, for the conversion, to use a weight that is unequal to one for an uni-prediction mode, and the conversion is performed based on the determining,

wherein the prediction of the current video block uses a final predictor corresponding to predictors scaled by the weight, and wherein a weighting factor set is selected at a block level or a coding unit level.

7. The method of claim 6 , wherein a weight index is signaled for at least one of an AMVP (Advanced Motion Vector Prediction) mode or an affine inter mode, and the weight index is inherited in at least one of a merge mode, an UMVE (Ultimate Motion Vector Expression) mode, an affine merge mode, or affine merge mode with offsets, and wherein the UMVE mode comprises a motion vector expression that includes a starting point, a motion magnitude and a motion direction for the current video block.

8. The method of claim 1 , wherein the performing of the conversion includes refining a motion compensation block computed during the conversion using an offset value from one of multiple offset sets.

9. The method of claim 8 , wherein an offset index is signaled at a block level or coding unit level, optionally, the offset index is inherited in at least one of a merge mode, an UMVE (Ultimate Motion Vector Expression) mode, an affine merge mode, or affine merge mode with offsets, and wherein the UMVE mode comprises a motion vector expression that includes a starting point, a motion magnitude and a motion direction for the current video block.

10. The method of claim 1 , wherein the rule specifies to select or derive a weighting factor for the GBi coding tool based on neighboring pixels of the current video block and corresponding reference neighboring pixels identified by motion vectors or integer part of the motion vectors of the current video block, wherein a signaling of a GBi index is omitted, and a field in the bitstream indicates whether to use a default equal weighting factor or a selected or derived weighting factor.

11. The method of claim 10 , wherein a weighting factor for the current video block is selected from a predefined weighting factor set to have a minimum difference between a result of averaging of the reference neighboring pixels from the two reference lists using the weighting factor and the neighboring pixels of the current video block, and

wherein GBi weighting values are derived by S0 and S1, S0 indicating a sum or an average of the neighboring pixels of the current video block and S1 indicating a sum or an average of neighboring pixels of a reference block.

12. The method of claim 1 , wherein the rule specifies to use local illumination compensation (LIC) parameters associated with the current video block for determining a weighting factor for the GBi coding tool, and

wherein the LIC parameters are derived to use a linear model of illumination changes in the current block during the conversion.

13. The method of claim 12 , wherein offsets to derive the LIC parameters are zero and only the weighting factor is used in the GBi coding tool, optionally both offsets and the weighting factor of the LIC parameters are used in the GBi coding tool,

wherein the rule specifies to inherit a GBi index for the current video block from that of a neighboring video block in a case that the current video block also uses at least one of a merge mode, an UMVE mode, an affine merge mode, or an affine merge mode with offsets, and

wherein the UMVE mode comprises a motion vector expression that includes a starting point, a motion magnitude and a motion direction for the current video block; optionally the rule specifies to derive additional LIC parameters for the conversion of the current video block in a case that an inherited neighboring weighting factor is same as the LIC parameters.

14. The method of claim 1 , wherein the conversion includes encoding the current video block into the bitstream.

15. The method of claim 1 , wherein the conversion includes decoding the current video block 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 conversion between a current video block of a video and a bitstream of the video, based on a rule that controls a parameter of a generalized bi-prediction (GBi) coding tool in which a prediction of the current video block uses a final predictor corresponding to a non-uniformly weighted sum of predictors from two reference lists;

determine, for the conversion, a mode of operation of the GBi coding tool, and the conversion is performed based on the determining,

wherein the rule specifies to reorder weighting factors for the GBi coding tool based on a difference between neighboring pixels of the current video block and a result of averaging of corresponding reference neighboring pixels identified by motion vectors or integer part of the motion vectors of the current video block using the weighting factor, wherein the reference neighboring pixels are from the two reference lists,

wherein a field in the bitstream corresponds to the mode of operation and a value of the field indicates whether the GBI coding tool is enabled or disabled for the current video block,

wherein the GBi coding tool is disabled for the current video block that uses a quantization parameter greater than a threshold.

17. A non-transitory computer-readable recording medium storing a bitstream of a video which is generated by a method performed by a video processing apparatus, wherein the method comprises:

generating the bitstream of the video based on a rule that controls a parameter of a generalized bi-prediction (GBi) coding tool in which a prediction of a current video block uses a final predictor corresponding to a non-uniformly weighted sum of predictors from two reference lists;

determining a mode of operation of the GBi coding tool, and the bitstream is generated based on the determining,

wherein the rule specifies to reorder weighting factors for the GBi coding tool based on a difference between neighboring pixels of the current video block and a result of averaging of corresponding reference neighboring pixels identified by motion vectors or integer part of the motion vectors of the current video block using the weighting factor, wherein the reference neighboring pixels are from the two reference lists,

wherein a field in the bitstream corresponds to the mode of operation and a value of the field indicates whether the GBI coding tool is enabled or disabled for the current video block,

wherein the GBi coding tool is disabled for the current video block that uses a quantization parameter greater than a threshold.