Encoder, decoder, encoding method, and decoding method
An encoder which includes circuitry and memory. Using the memory, the circuitry generates a list which includes candidates for a first motion vector for a first partition. The list has a maximum list size and an order of the candidates, and at least one of the maximum list size or the order of the candidates is dependent on at least one of a partition size or a partition shape of the first partition. The circuitry selects the first motion vector from the candidates included in the list; encodes an index indicating the first motion vector among the candidates in the list into the bitstream based on the maximum list size; and generates the predicted image for the first partition using the first motion vector.
1. A decoder which decodes a video from a bitstream using a predicted image, the decoder comprising:
circuitry; and
memory,
wherein, using the memory, the circuitry:
derives a maximum number based on a partition shape of a first partition in the video, the maximum number being a first number when the partition shape is a square or a rectangle, and the maximum number being a second number different from the first number when the partition shape is a non-square and a non-rectangle;
generates a motion vector candidate list including a plurality of candidates equal to the derived maximum number, the plurality of candidates being for a first motion vector for the first partition, and the plurality of candidates including a candidate which is derived from a second motion vector of a second partition in the video different from the first partition;
decodes an index indicating the first motion vector among the plurality of candidates in the motion vector candidate list from the bitstream based on the maximum number,
selects the first motion vector from the plurality of candidates in the motion vector candidate list using the index; and
generates the predicted image for the first partition using the first motion vector.
2. A decoding method of decoding a video from a bitstream using a predicted image, the decoding method comprising:
deriving a maximum number based on a partition shape of a first partition in the video, the maximum number being a first number when the partition shape is a square or a rectangle, and the maximum number being a second number different from the first number when the partition shape is a non-square and a non-rectangle;
generating a motion vector candidate list including a plurality of candidates equal to the derived maximum number, the plurality of candidates being for a first motion vector for the first partition, and the plurality of candidates including a candidate which is derived from a second motion vector of a second partition in the video different from the first partition;
decoding an index indicating the first motion vector among the plurality of candidates in the motion vector candidate list from the bitstream based on the maximum number;
selecting the first motion vector from the plurality of candidates in the motion vector candidate list using the index; and
generating the predicted image for the first partition using the first motion vector.
3. A non-transitory computer readable medium storing a bitstream, the bitstream including syntax information according to which a computer performs a decoding process of decoding a video from the bitstream using a predicted image, the decoding process comprising:
deriving a maximum number based on a partition shape of a first partition in the video, the maximum number being a first number when the partition shape is a square or a rectangle, and the maximum number being a second number different from the first number when the partition shape is a non-square and a non-rectangle;
generating a motion vector candidate list including a plurality of candidates equal to the derived maximum number, the plurality of candidates being for a first motion vector for the first partition, and the plurality of candidates including a candidate which is derived from a second motion vector of a second partition in the video different from the first partition;
decoding an index indicating the first motion vector among the plurality of candidates in the motion vector candidate list from the bitstream based on the maximum number;
selecting the first motion vector from the plurality of candidates in the motion vector candidate list using the index; and
generating the predicted image for the first partition using the first motion vector.