Signaling a maximum number of merge motion vector prediction candidates for inter prediction
A method for video encoding includes generating a first syntax element indicating a maximum number of merge motion vector prediction (MVP) candidates, and encoding the first syntax element at a first coding level in a coded video bitstream, the first coding level being higher than a picture parameter set (PPS) level. The video encoding method also includes transmitting the coded video bitstream including the first syntax element. The maximum number of merge MVP candidates indicated by the first syntax element is applicable for reconstructing coding blocks of the coded video bitstream associated with the first coding level.
1. A method for video encoding, comprising:
generating, by processing circuitry of a video encoder, a set of syntax elements defining maximum numbers of merge candidates for a picture sequence in a coded video bitstream, the set of syntax elements including at least one of a first syntax element indicating a maximum number of merge motion vector prediction (MVP) candidates, a second syntax element indicating a maximum number of geometric merge mode merge candidates, a third syntax element indicating a maximum number of subblock-based merge MVP candidates, and a fourth syntax element indicating a maximum number of intra block copy (IBC) merge block vector prediction (BVP) candidates;
encoding the set of syntax elements defining the maximum numbers of merge candidates in a sequence parameter set (SPS) in the coded video bitstream; and
transmitting the coded video bitstream including the set of syntax elements,
wherein coding blocks of the picture sequence in the coded video bitstream are coded using the maximum numbers of merge candidates indicated by the set of syntax elements.
2. The method of claim 1 , further comprising:
including, in the set of syntax elements in the coded video bitstream, the second syntax element, the third syntax element, and the fourth syntax element.
3. The method of claim 1 , further comprising:
including, in the coded video bitstream and based on a condition being satisfied, a fifth syntax element at a second coding level, the fifth syntax element indicating a maximum number of merge candidates for one of a geometric merge mode, a subblock-based merge mode, and an intra block copy (IBC) merge mode.
4. The method of claim 3 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the geometric merge mode;
the maximum number of merge candidates is a maximum number of geometric merge mode merge candidates;
the fifth syntax element indicates the maximum number of geometric merge mode merge candidates; and
the condition is the geometric merge mode being enabled at the second coding level and the maximum number of merge MVP candidates being larger than or equal to 3.
5. The method of claim 3 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the subblock-based merge mode;
the maximum number of merge candidates is a maximum number of subblock-based merge MVP candidates;
the fifth syntax element indicates the maximum number of subblock-based merge MVP candidates; and
the condition is an affine mode being enabled at the second coding level.
6. The method of claim 3 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the IBC merge mode;
the maximum number of merge candidates is a maximum number of IBC merge block vector prediction (BVP) candidates;
the fifth syntax element indicates the maximum number of IBC merge BVP candidates; and
the condition is the IBC merge mode being enabled at the second coding level.
7. The method of claim 3 , wherein the second coding level is a sequence level.
8. The method of claim 5 , wherein the fifth syntax element is modified at a coding level that is lower than the second coding level.
9. The method of claim 8 , wherein the second coding level is a sequence level, and the coding level that is lower than the second coding level is a picture level.
10. An apparatus for video decoding, comprising:
processing circuitry configured to
decode, from a coded video bitstream, a sequence parameter set (SPS) including a set of syntax elements defining maximum numbers of merge candidates for a picture sequence in the coded video bitstream, the set of syntax elements including at least one of a first syntax element indicating a maximum number of merge motion vector prediction (MVP) candidates, a second syntax element indicating a maximum number of geometric merge mode merge candidates, a third syntax element indicating a maximum number of subblock-based merge MVP candidates, and a fourth syntax element indicating a maximum number of intra block copy (IBC) merge block vector prediction (BVP) candidates; and
decode coding blocks of the picture sequence in the coded video bitstream using the maximum numbers of merge candidates indicated by the set of syntax elements.
11. The apparatus of claim 10 , wherein the processing circuitry is further configured to:
decode, from the set of syntax elements in the coded video bitstream, the second syntax element, the third syntax element, and the fourth syntax element.
12. The apparatus of claim 10 , wherein the processing circuitry is further configured to:
decode, from the coded video bitstream and based on a condition being satisfied, a fifth syntax element signaled at a second coding level, the fifth syntax element indicating a maximum number of merge candidates for one of a geometric merge mode, a subblock-based merge mode, and an intra block copy (IBC) merge mode.
13. The apparatus of claim 12 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the geometric merge mode;
the maximum number of merge candidates is a maximum number of geometric merge mode merge candidates;
the fifth syntax element indicates the maximum number of geometric merge mode merge candidates; and
the condition is the geometric merge mode being enabled at the second coding level and the maximum number of merge MVP candidates being larger than or equal to 3.
14. The apparatus of claim 12 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the subblock-based merge mode;
the maximum number of merge candidates is a maximum number of subblock-based merge MVP candidates;
the fifth syntax element indicates the maximum number of subblock-based merge MVP candidates; and
the condition is an affine mode being enabled at the second coding level.
15. The apparatus of claim 12 , wherein
the one of the geometric merge mode, the subblock-based merge mode, and the IBC merge mode is the IBC merge mode;
the maximum number of merge candidates is a maximum number of IBC merge block vector prediction (BVP) candidates;
the fifth syntax element indicates the maximum number of IBC merge BVP candidates; and
the condition is the IBC merge mode being enabled at the second coding level.
16. The apparatus of claim 12 , wherein the second coding level is a sequence level.
17. The apparatus of claim 14 , wherein the fifth syntax element is modified at a coding level that is lower than the second coding level.
18. A method of processing visual media data, the method comprising:
processing a bitstream of the visual media data according to a format rule, wherein
the bitstream includes a set of syntax elements defining maximum numbers of merge candidates for a picture sequence in the bitstream, the set of syntax elements including at least one of a first syntax element indicating a maximum number of merge motion vector prediction (MVP) candidates, a second syntax element indicating a maximum number of geometric merge mode merge candidates, a third syntax element indicating a maximum number of subblock-based merge MVP candidates, and a fourth syntax element indicating a maximum number of intra block copy (IBC) merge block vector prediction (BVP) candidates;
the format rule specifies that
the set of syntax elements defining the maximum numbers of merge candidates is encoded in a sequence parameter set (SPS) in the bitstream;
the bitstream including the set of syntax elements is transmitted; and
coding blocks of the picture sequence in the bitstream are coded using the maximum numbers of merge candidates indicated by the set of syntax elements.