Entropy coding of motion vector differences
An entropy decoder is configured to, for horizontal and vertical components of motion vector differences, derive a truncated unary code from the data stream using context-adaptive binary entropy decoding with exactly one context per bin position of the truncated unary code, which is common for horizontal and vertical components of the motion vector differences, and an Exp-Golomb code using a constant equi-probability bypass mode to obtain the binarizations of the motion vector differences. A desymbolizer is configured to debinarize the binarizations of the motion vector difference syntax elements to obtain integer values of the horizontal and vertical components of the motion vector differences. A reconstructor is configured to reconstruct a video based on the integer values of the horizontal and vertical components of the motion vector differences.
1. A decoder for decoding a video encoded in a data stream, comprising:
a desymbolizer configured to:
debinarize a binarization of a motion vector difference, the motion vector difference representing a prediction error with respect to a motion vector used in motion-compensated prediction coding of a coding block of the video,
the binarization of the motion vector difference including (a) a prefix bin string including fixed length binary codes based on a cutoff value equal to two, wherein one of the fixed length binary codes represents whether an absolute value of the motion vector difference is greater than zero, and (b) a suffix bin string including an Exp-Golomb code having a fixed order set to one; and
a reconstructor configured to:
determine a set of motion vector predictors, wherein at least one motion vector predictor of the set is based on a motion vector of a neighbor of a coding block associated with the motion vector difference,
obtain, from the data stream, an index indicating a specific motion vector predictor of the set of motion vector predictors,
reconstruct the motion vector based on a debinarized value of the motion vector difference and the specific motion vector predictor, and
reconstruct the video based on the reconstructed motion vector.
2. The decoder of claim 1 , wherein the data stream comprises at least a portion associated with color samples of the video.
3. The decoder of claim 1 , wherein the data stream comprises at least a portion associated with depth values related to a depth map associated with the video.
4. The decoder of claim 1 , further comprising an entropy decoder configured to decode the fixed length binary codes using binary arithmetic decoding.
5. The decoder of claim 4 , wherein the entropy decoder is configured to decode the fixed length binary codes prior to decoding the Exp-Golomb code.
6. The decoder of claim 1 , wherein the neighbor is a spatial neighbor of the coding block.
7. The decoder of claim 1 , wherein another of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than one.
8. The decoder of claim 1 , wherein the reconstructor is configured to reconstruct the video using the motion-compensated prediction coding by applying the reconstructed motion vector at a spatial granularity defined by a sub-division of the video's pictures in blocks, wherein the reconstructor uses merging syntax elements present in the data stream so as to group the blocks into merge groups and apply the reconstructed motion vector in units of merge groups.
9. The decoder of claim 8 , wherein the reconstructor is configured to derive the sub-division of the video's pictures in blocks from a portion of the data stream excluding the merging syntax elements.
10. A method for decoding a video encoded in a data stream, comprising:
debinarizing a binarization of a motion vector difference, the motion vector difference representing a prediction error with respect to a motion vector used in motion-compensated prediction coding of a coding block of the video, the binarization of the motion vector difference including (a) a prefix bin string including fixed length binary codes based on a cutoff value equal to two, wherein one of the fixed length binary codes represents whether an absolute value of the motion vector difference is greater than zero, and (b) a suffix bin string including an Exp-Golomb code having a fixed order set to one;
determining a set of motion vector predictors, wherein at least one motion vector predictor of the set is based on a motion vector of a neighbor of a coding block associated with the motion vector difference;
obtaining, from the data stream, an index indicating a specific motion vector predictor of the set of motion vector predictors;
reconstructing the motion vector based on a debinarized value of the motion vector difference and the specific motion vector predictor; and
reconstructing the video based on the reconstructed motion vector.
11. The method of claim 10 , wherein the data stream comprises at least a portion associated with color samples of the video.
12. The method of claim 10 , wherein the data stream comprises at least a portion associated with depth values related to a depth map associated with the video.
13. The method of claim 10 , further comprising decoding the fixed length binary codes using binary arithmetic decoding.
14. The method of claim 13 , wherein the decoding of the fixed length binary codes is performed prior to decoding of the Exp-Golomb code.
15. The method of claim 10 , wherein the neighbor is a spatial neighbor of the coding block.
16. The method of claim 10 , wherein another of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than one.
17. The method of claim 10 , wherein the reconstructing comprises reconstructing the video using the motion-compensated prediction coding by applying the reconstructed motion vector at a spatial granularity defined by a sub-division of the video's pictures in blocks, wherein the reconstructing uses merging syntax elements present in the data stream so as to group the blocks into merge groups and apply the reconstructed motion vector in units of merge groups.
18. The method of claim 17 , further comprising deriving the sub-division of the video's pictures in blocks from a portion of the data stream excluding the merging syntax elements.
19. An encoder for encoding a video into a data stream, comprising:
an inserter configured to:
insert, into the data stream, an encoded binarized value of a motion vector difference, the motion vector difference representing a prediction error with respect to a motion vector used in motion-compensated prediction coding of a coding block of the video, wherein the binarized value of the motion vector difference includes (a) a prefix bin string including a fixed length binary codes based on a cutoff value equal to two, wherein one of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than zero, and (b) a suffix bin string including an Exp-Golomb code having a fixed order set to one, and
insert, into the data stream, an index indicating a specific motion vector predictor of a set of motion vector predictors, wherein at least one of the set of motion vector predictors is determined based on a motion vector of a neighbor of the coding block, and the motion vector difference is determined based on the motion vector and the specific motion vector predictor.
20. The encoder of claim 19 , wherein the data stream comprises at least a portion associated with color samples of the video.
21. The encoder of claim 19 , wherein the data stream comprises at least a portion associated with depth values related to a depth map associated with the video.
22. The encoder of claim 19 , further comprising an entropy encoder configured to encode the fixed length binary codes using binary arithmetic coding.
23. The encoder of claim 22 , wherein the entropy encoder is configured to encode the fixed length binary codes prior to encoding the Exp-Golomb code.
24. The encoder of claim 19 , wherein the neighbor is a spatial neighbor of the coding block.
25. The encoder of claim 19 , wherein another of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than one.
26. A non-transitory computer-readable medium for storing data associated with a video, comprising:
a data stream stored in the non-transitory computer-readable medium, the data stream comprising an encoded binarized value of a motion vector difference, the motion vector difference representing a prediction error with respect to a motion vector used in motion-compensated prediction coding of a coding block of the video, wherein the binarized value of the motion vector difference includes (a) a prefix bin string including a fixed length binary codes based on a cutoff value equal to two, wherein one of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than zero, and (b) a suffix bin string including an Exp-Golomb code having a fixed order set to one, and
the data stream comprising an index indicating a specific motion vector predictor of a set of motion vector predictors, wherein at least one of the set of motion vector predictors is determined based on a motion vector of a neighbor of the coding block, and the motion vector difference is determined based on the motion vector and the specific motion vector predictor.
27. The non-transitory computer-readable medium of claim 26 , wherein the data stream comprises at least a portion associated with color samples of the video.
28. The non-transitory computer-readable medium of claim 26 , wherein the data stream comprises at least a portion associated with depth values related to a depth map associated with the video.
29. The non-transitory computer-readable medium of claim 26 , wherein the fixed length binary codes is entropy encoded into the data stream using binary arithmetic coding.
30. The non-transitory computer-readable medium of claim 26 , wherein another of the fixed length binary codes indicates whether an absolute value of the motion vector difference is greater than one.