Methods and apparatuses for cross-component prediction

1. A method of video processing, comprising:

determining, for a conversion between a current video block of a video that is a chroma block coded with a cross-component prediction mode and a bitstream of the video, an offset parameter of a cross-component prediction model that is based on a derived sample value from two or more neighbor samples of the current video block, wherein the derived sample value is derived from neighbor reconstructed samples of the current video block and the derived sample value is a derived chroma value or a derived luma value, wherein the two or more neighbor samples include all neighbor samples of the current video block, or the two or more neighbor samples are from a group of neighbor samples, wherein the group includes a portion of all neighbor samples of the current video block, and wherein all neighbor samples are classified into different groups based on at least one of sample values, or any functions on sample values, or variances, or distributions, or activities, or diversities, wherein all neighbor samples refer to a whole set of samples within a region close to the current video block; and

performing the conversion based on the determining.

2. The method of claim 1 , wherein the two or more neighbor samples refer to neighbor luma samples and/or neighbor chroma samples located at top rows and/or left columns outside the current video block.

3. The method of claim 1 , wherein the derived sample value is based on at least one of an average chroma value, a midrange chroma value, a median chroma value, an average luma value, a midrange luma value, or a median luma value of the two or more neighbor samples.

4. The method of claim 1 , wherein determining the offset parameter comprises determining the offset parameter based on the derived chroma value or based on the derived chroma value and a parameter dependent on the derived luma value.

5. The method of claim 1 , wherein the derived sample value is a sum of a minimum value, a maximum value, and an offset factor, wherein the sum is right-shifted by a shift factor.

6. The method of claim 5 , wherein:

the minimum value is a minimum value associated with the two or more neighbor samples of the current video block; and/or

the maximum value is a maximum value associated with the two or more neighbor samples of the current video block.

7. The method of claim 5 , wherein:

the minimum value is an average of N smallest neighbor samples of two or more neighbor samples of the current video block, wherein N is a constant; and/or

the maximum value is an average of M largest neighbor samples of two or more neighbor samples of the current video block, wherein M is a constant.

8. The method of claim 1 , wherein the derived sample value is a sum of a parameter S and an offset factor, wherein the sum is right-shifted by a shift factor.

9. The method of claim 8 , wherein the shift factor is dependent on a number of the neighbor reconstructed samples.

10. The method of claim 8 , wherein the parameter S is a parameter sum of L neighbor samples of the current video block, wherein L equals to a×blockWidth, b×blockHeight, or c×(blockWidth+blockHeight), wherein a, b, and c are integers.

11. The method of claim 1 , wherein determining the offset parameter further comprises determining the offset parameter based on one or more models.

12. The method of claim 1 , wherein determining the offset parameter further comprises determining the offset parameter during a cross-component linear model left-top (CCLM-LT) mode, a cross-component linear model left (CCLM-L) mode, or a cross-component linear model top (CCLM-T) mode.

13. The method of claim 1 , wherein the offset parameter is a difference between a derived chroma and a product of a and a derived luma value, wherein a is scaling parameter applying to a luma reconstructed value and the difference is right-shifted by a shift factor.

14. The method of claim 1 , wherein the two or more neighbor samples are adjacent to the current video block, wherein the current video block includes at least one of a chroma block or a luma block; or

the two or more neighbor samples are non-adjacent to the current video block, wherein the current video block includes at least one of a chroma block or a luma block.

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

16. The method of claim 1 , wherein the conversion includes decoding the current video block from the bitstream.

17. The method of claim 1 , wherein the conversion includes generating the bitstream from the current video block; and the method further comprises:

storing the bitstream in a non-transitory computer-readable recording medium.

18. 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:

determine, for a conversion between a current video block of a video that is a chroma block coded with a cross-component prediction mode and a bitstream of the video, an offset parameter of a cross-component prediction model that is based on a derived sample value from two or more neighbor samples of the current video block, wherein the derived sample value is derived from neighbor reconstructed samples of the current video block and the derived sample value is a derived chroma value or a derived luma value, wherein the two or more neighbor samples include all neighbor samples of the current video block, or the two or more neighbor samples are from a group of neighbor samples, wherein the group includes a portion of all neighbor samples of the current video block, and wherein all neighbor samples are classified into different groups based on at least one of sample values, or any functions on sample values, or variances, or distributions, or activities, or diversities, wherein all neighbor samples refer to a whole set of samples within a region close to the current video block; and

perform the conversion based on the determining.

19. 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:

determining, for a conversion between a current video block of a video that is a chroma block coded with a cross-component prediction mode and a bitstream of the video, an offset parameter of a cross-component prediction model that is based on a derived sample value from two or more neighbor samples of the current video block, wherein the derived sample value is derived from neighbor reconstructed samples of the current video block and the derived sample value is a derived chroma value or a derived luma value, wherein the two or more neighbor samples include all neighbor samples of the current video block, or the two or more neighbor samples are from a group of neighbor samples, wherein the group includes a portion of all neighbor samples of the current video block, and wherein all neighbor samples are classified into different groups based on at least one of sample values, or any functions on sample values, or variances, or distributions, or activities, or diversities, wherein all neighbor samples refer to a whole set of samples within a region close to the current video block; and

generating the bitstream from the current video block based on the determining.

20. A non-transitory computer-readable storage medium storing instructions that cause a processor to:

determine, for a conversion between a current video block of a video that is a chroma block coded with a cross-component prediction mode and a bitstream of the video, an offset parameter of a cross-component prediction model that is based on a derived sample value from two or more neighbor samples of the current video block, wherein the derived sample value is derived from neighbor reconstructed samples of the current video block and the derived sample value is a derived chroma value or a derived luma value, wherein the two or more neighbor samples include all neighbor samples of the current video block, or the two or more neighbor samples are from a group of neighbor samples, wherein the group includes a portion of all neighbor samples of the current video block, and wherein all neighbor samples are classified into different groups based on at least one of sample values, or any functions on sample values, or variances, or distributions, or activities, or diversities, wherein all neighbor samples refer to a whole set of samples within a region close to the current video block; and

perform the conversion based on the determining.