Efficient transform signaling for small blocks
A size of a transform block is identified. A transform type for the transform block is identified. The transform type includes a horizontal transform type and a vertical transform type. Identifying the transform type includes determining whether the size of the transform block is below a predefined block size; and, in response to determining that the size is below the predefined block size, selecting a default transform type for each of the horizontal transform type and the vertical transform type. The transform type is then applied to the transform block. The default transform type can be the discrete cosine transform (DCT).
1 . A method for coding a transform block, comprising:
identifying a size of the transform block;
identifying a transform type for the transform block, wherein the transform type comprises a horizontal transform type and a vertical transform type, and wherein identifying the transform type comprises:
determining whether the size of the transform block is below a predefined block size;
in response to determining that the size is below the predefined block size, selecting a default transform type for each of the horizontal transform type and the vertical transform type; and
in response to determining that the size is not below the predefined block size and that one of a width or a height of the transform block is equal to a minimum threshold transform dimension size, selecting the default transform type for the one of the width or the height of the transform block that is equal to the minimum threshold transform dimension size; and
applying the transform type to the transform block.
2 . The method of claim 1 , wherein the threshold size is 4.
3 . The method of claim 1 , wherein identifying the transform type further comprises:
in response to determining that the size is not below the predefined block size:
identifying a prediction mode associated with the transform block;
identifying a subset of available transform types based on the prediction mode; and
coding an index into the subset of the transform type.
4 . The method of claim 1 , further comprising:
determining based on the size of the transform block that the transform block is not split into sub-blocks.
5 . The method of claim 4 , further comprising:
inferring a syntax element indicating whether the transform block is split into the sub-blocks based on the size.
6 . The method of claim 1 , wherein the default transform type is a discrete cosine transform.
7 . A device, comprising:
a processor configured to execute instructions for coding a transform block, the instructions comprise to:
identify a size of the transform block;
identify a transform type for the transform block, wherein the transform type comprises a horizontal transform type and a vertical transform type, and wherein identifying the transform type comprises:
determine whether the size of the transform block is below a predefined block size;
in response to determining that the size is below the predefined block size, select a default transform type for each of the horizontal transform type and the vertical transform type;
in response to determining that the size is not below the predefined block size and that one of a width or a height of the transform block is equal to a minimum threshold transform dimension size, select the default transform type for the one of the width or the height of the transform block that is equal to the minimum threshold transform dimension size; and
apply the transform type to the transform block.
8 . The device of claim 7 , wherein the threshold size is 4.
9 . The device of claim 7 , wherein the instructions to identify the transform type further comprise instructions executable by the processor to:
in response to determining that the size is not below the predefined block size:
identify a prediction mode associated with the transform block;
identify a subset of available transform types based on the prediction mode; and
code an index into the subset of the transform type.
10 . The device of claim 7 , wherein the processor is further configured to execute instructions to:
determine based on the size of the transform block that the transform block is not split into sub-blocks.
11 . The device of claim 10 , wherein the processor is further configured to execute instructions to:
infer a syntax element indicating whether the transform block is split into the sub-blocks based on the size.
12 . The device of claim 10 , wherein the default transform type is a discrete cosine transform.
13 . A non-transitory computer-readable storage medium having stored thereon an encoded bitstream, wherein the encoded bitstream is configured for decoding by operations comprising:
identifying a size of a transform block;
identifying a transform type for the transform block, wherein the transform type comprises a horizontal transform type and a vertical transform type, and wherein identifying the transform type comprises:
determining whether the size of the transform block is below a predefined block size;
in response to determining that the size is below the predefined block size, selecting a default transform type for each of the horizontal transform type and the vertical transform type;
in response to determining that the size is not below the predefined block size and that one of a width or a height of the transform block is equal to a minimum threshold transform dimension size, selecting the default transform type for the one of the width of the height of the transform block that is equal to the minimum threshold transform dimension size; and
applying the transform type to the transform block.
14 . The non-transitory computer-readable storage medium of claim 13 , wherein the default transform type is a discrete cosine transform and the threshold size is 4.
15 . The non-transitory computer-readable storage medium of claim 13 , wherein identifying the transform type further comprises:
in response to determining that the size is not below the predefined block size:
identifying a prediction mode associated with the transform block;
identifying a subset of available transform types based on the prediction mode; and
coding an index into the subset of the transform type.
16 . The non-transitory computer-readable storage medium of claim 13 , the operations further comprising:
determining based on the size of the transform block that the transform block is not split into sub-blocks.
17 . The non-transitory computer-readable storage medium of claim 16 , the operations further comprising:
inferring a syntax element indicating whether the transform block is split into the sub-blocks based on the size.
18 . The method of claim 3 , wherein the prediction mode is an optical flow mode that refines motion vectors subblock-wise.
19 . The method of claim 3 , wherein the prediction mode is a compound prediction mode.
20 . The method of claim 3 , wherein identifying the subset of available transform types based on the prediction mode comprises:
in response to the size of the transform block being 8×8 and the prediction mode being an optical flow mode, identifying a subset comprising only a DCT_DCT transform type.