Signaling for video block partition mode

1. A method of processing video data, further comprising:

determining a valid splitting type for a current video block from candidates of splitting types;

determining whether or how to signal a splitting type to be used for the current video block based on a determination of the valid splitting type; and

performing a conversion between the current video block and a bitstream of the current video block in accordance with results of the determining,

wherein the candidates of the splitting types at least include unsymmetrical quad-trees (UQTs) splitting, and

wherein, in the unsymmetrical quad-Trees (UQTs) splitting, the current video block is unsymmetrically split into four partitions, and at least one of the four partitions has a block size different from block sizes of other partitions of the four partitions,

wherein, when only one kind of UQT splitting is valid and UQT splitting is signaled to be used, refraining from signaling further information to indicate which kind of UQT splitting is used,

wherein, when only two kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a flag to indicate which kind of UQT splitting is used, and

wherein, when only three kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a message to indicate which kind of UQT splitting is used.

2. The method of claim 1 , wherein candidates of the splitting types to be signaled include binary tree (BT) splitting, triple tree (TT) splitting or unsymmetrical quad-trees (UQTs) splitting, and the candidates are all vertical splitting or horizontal splitting depending on previously signaled or derived information during the conversion.

3. The method of claim 2 , wherein determining the splitting type used for the current video block is based on bins of a bin string coded by arithmetic coding.

4. The method of claim 3 , wherein partial bins of the bin string are coded with contexts and other bins are coded with bypass mode without contexts; or

wherein all bins of the bin string are coded with contexts; or

wherein all bins of the bin string are coded with bypass mode, and

wherein one or multiple contexts are utilized for a bin coded with contexts, and

wherein a context of the contexts is based on at least one of: a position or index of the bin, splitting of neighboring blocks of the current video block, a current split depth of the current video block, a split depth of neighboring blocks, coding modes of neighboring blocks, a width or height of neighboring blocks, a width or height of the current video block, a slice type, a picture type, a tile group type, a color component, or statistics results of splitting types from previously coded video blocks.

5. The method of claim 1 ,

wherein, when there is only one splitting type among BT splitting, TT splitting, and UQT splitting that is valid, refraining from signaling any binarized code to indicate the splitting type used for the current video block; and

wherein, when there are only two splitting types among BT splitting, TT splitting, and UQT splitting that are valid, signaling a flag to indicate which one of two valid splitting types is used, wherein a code to indicate the splitting type to be used is binarized as a truncated unary code, and wherein a maximum value of the truncated unary code is N−1, where N is an integer and represents a number of valid splitting types.

6. The method of claim 1 , wherein there is no codeword to represent an invalid splitting type in the bitstream, and wherein the invalid splitting type is skipped when building a codeword table.

7. The method of claim 1 , wherein indications of candidates of splitting types are signaled in the bitstream after signaling at least one of indications of splitting or non-splitting and indications of quad tree (QT) splitting, wherein the candidates of splitting types include: extended quad-tree (EQT) splitting, binary tree (BT) splitting and unsymmetrical quad-trees (UQT) splitting.

8. The method of claim 7 , wherein in UQT splitting in which the current video block is split into four partitions including a first partition having a dimension of W1×H1, a second partition having a dimension of W2×H2, a third partition having a dimension of W3×H3, and a fourth partition having a dimension of W4×H4, wherein W1, W2, W3, W4, H1, H2, H3, and H4 are integers.

9. The method of claim 8 , wherein a first flag is signaled to indicate whether the current video block is split with a first set comprising UQT splitting and BT splitting or a second set comprising UQT splitting and EQT splitting, wherein a UQT splitting type in the first set and the second set are different, and

wherein, when the first flag indicates that the current video block is split with the first set, a second flag is further signaled to indicate whether a certain kind of UQT splitting is applied or a certain kind of BT splitting is applied on the current video block, and wherein the second flag is signaled to indicate whether UQT2-V/UQT2-H splitting is applied, or BT-V/BT-H splitting is applied on the current video block.

10. The method of claim 1 , further comprising:

determining, for a conversion between a current video block and a bitstream of the current video block, interpretations of a signaled representation of unsymmetrical quad-tree (UQT) splitting to be used for splitting the current video block, based on a slice or a picture containing the current video block, wherein the signaled representation of UQT splitting in an I-slice/I-picture is different to that in a P-slice/P-picture or a B-slice/B-picture.

11. The method of claim 1 , wherein how to signal the splitting type depends on a picture/slice type containing the current video block.

12. The method of claim 11 , wherein a UQT flag to indicate whether UQT splitting is applied is signaled before a BT/EQT flag to indicate whether BT/EQT splitting is applied in an I picture/slice, and the UQT flag to indicate whether UQT splitting is applied is signaled after the BT/EQT flag to indicate whether BT/EQT splitting is applied in a P picture/slice or B picture/slice; or alternatively,

wherein a UQT flag to indicate whether UQT splitting is applied is signaled after a BT/EQT flag to indicate whether BT/EQT splitting is applied in an I picture/slice, and the UQT flag to indicate whether UQT splitting is applied is signaled before the BT/EQT flag to indicate whether BT/EQT splitting is applied in a P picture/slice or B picture/slice.

13. The method of claim 12 , wherein the UQT flag to indicate whether UQT splitting is applied is coded by arithmetic coding with one or more context models.

14. The method of claim 13 , wherein the one or more context models are selected based on the picture/slice type, wherein the picture/slice type includes I picture/slice and P/B picture/slice; or

wherein the one or more context models are selected based on whether there is one or multiple neighboring blocks; or

wherein the one or more context models are selected based on dimensions of one or multiple neighboring blocks.

15. The method of claim 1 , wherein a flag is signaled to indicate a splitting direction of more than one splitting types including UQT splitting, and wherein the flag is signaled to indicate whether vertical or horizontal splitting is applied for UQT splitting, EQT splitting and BT splitting.

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

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

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 a valid splitting type for a current video block from candidates of splitting types;

determine whether or how to signal a splitting type to be used for the current video block based on a determination of the valid splitting type; and

perform a conversion between the current video block and a bitstream of the current video block in accordance with results of the determining,

wherein the candidates of the splitting types at least include unsymmetrical quad-trees (UQTs) splitting, and

wherein in the unsymmetrical quad-Trees (UQTs) splitting, the current video block is unsymmetrically split into four partitions, and at least one of the four partitions has a block size different from block sizes of other partitions of the four partitions,

wherein, when only one kind of UQT splitting is valid and UQT splitting is signaled to be used, refraining from signaling further information to indicate which kind of UQT splitting is used,

wherein, when only two kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a flag to indicate which kind of UQT splitting is used, and

wherein, when only three kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a message to indicate which kind of UQT splitting is used.

19. A method of storing a bitstream of a video, comprising:

determining a valid splitting type for a current video block from candidates of splitting types;

determining whether or how to signal a splitting type to be used for the current video block based on a determination of the valid splitting type;

generating the bitstream in accordance with results of the determining; and

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

wherein the candidates of the splitting types at least include unsymmetrical quad-trees (UQTs) splitting, and

wherein in the unsymmetrical quad-Trees (UQTs) splitting, the current video block is unsymmetrically split into four partitions, and at least one of the four partitions has a block size different from block sizes of other partitions of the four partitions,

wherein, when only one kind of UQT splitting is valid and UQT splitting is signaled to be used, refraining from signaling further information to indicate which kind of UQT splitting is used,

wherein, when only two kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a flag to indicate which kind of UQT splitting is used, and

wherein, when only three kinds of UQT splitting are valid and UQT splitting is signaled to be used, signaling a message to indicate which kind of UQT splitting is used.