Position-based coefficients scaling
AssigneesBeijing Bytedance Network Technology Co., Ltd., Bytedance Inc.
InventorsLi Zhang, Kai Zhang, Hongbin Liu, Jizheng Xu, Yue Wang
Life of this patent
Estimating the term…
Abstract
A video processing method includes determining, according to a rule, whether to apply a scaling matrix based on whether a secondary transform matrix is applied to a portion of a video block of a video, wherein the scaling matrix is used to scale at least some coefficients of the video block, and wherein the secondary transform matrix is used to transform at least some residual coefficients of the portion of the video block during the conversion; and performing a conversion between the video block of the video and a bitstream representation of the video using the selected scaling matrix.
Claims14
A method of processing video data, comprising:
- determining, during a conversion between a video transform block of a video with a K×L transform block size and a bitstream of the video, that in response to the K×L video transform block containing more than 32×32 positions, only matrix elements of a scaling matrix within a top-left 32×32 region of the K×L video transform block are kept and remaining matrix elements of the scaling matrix are zeroing-out, and
- performing the conversion using the scaling matrix, wherein the scaling matrix is used to scale at least some coefficients of the video transform block,
- wherein a secondary transform matrix is applied to an S×T top left portion of the video transform block and a default scaling matrix or a scaling matrix explicitly included in the bitstream is applied to the video transform block as the scaling matrix, wherein S, T, K and L are integers and K is greater than or equal to S, L is greater than or equal to T, and wherein matrix element values of the default scaling matrix are equal to 16.
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, during a conversion between a video transform block of a video with a K×L transform block size and a bitstream of the video, that in response to the K×L video transform block containing more than 32×32 positions, only matrix elements of a scaling matrix within a top-left 32×32 region of the K×L video transform block are kept and remaining matrix elements of the scaling matrix are zeroing-out, and
- perform the conversion using the scaling matrix, wherein the scaling matrix is used to scale at least some coefficients of the video transform block,
- wherein a secondary transform matrix is applied to an S×T top left portion of the video transform block and a default scaling matrix or a scaling matrix explicitly included in the bitstream is applied to the video transform block as the scaling matrix, wherein S, T, K and L are integers and K is greater than or equal to S, L is greater than or equal to T, and wherein matrix element values of the default scaling matrix are equal to 16.
A non-transitory computer-readable storage medium storing instructions that cause a processor to:
- determine, during a conversion between a video transform block of a video with a K×L transform block size and a bitstream of the video, that in response to the K×L video transform block containing more than 32×32 positions, only matrix elements of a scaling matrix within a top-left 32×32 region of the K×L video transform block are kept and remaining matrix elements of the scaling matrix are zeroing-out, and
- perform the conversion using the scaling matrix, wherein the scaling matrix is used to scale at least some coefficients of the video transform block,
- wherein a secondary transform matrix is applied to an S×T top left portion of the video transform block and a default scaling matrix or a scaling matrix explicitly included in the bitstream is applied to the video transform block as the scaling matrix, wherein S, T, K and L are integers and K is greater than or equal to S, L is greater than or equal to T, and wherein matrix element values of the default scaling matrix are equal to 16.
A non-transitory computer-readable recording medium storing a bitstream of a video which is generated by a method performed by a video processing apparatus, wherein the method comprises:
- determining, for a video transform block of a video with a K×L transform block size, that in response to the K×L video transform block containing more than 32×32 positions, only matrix elements of a scaling matrix within a top-left 32×32 region of the K×L video transform block are kept and remaining matrix elements of the scaling matrix are zeroing-out, and
- generating the bitstream using the scaling matrix, wherein the scaling matrix is used to scale at least some coefficients of the video transform block,
- wherein a secondary transform matrix is applied to an S×T top left portion of the video transform block and a default scaling matrix or a scaling matrix explicitly included in the bitstream is applied to the video transform block as the scaling matrix, wherein S, T, K and L are integers and K is greater than or equal to S, L is greater than or equal to T, and wherein matrix element values of the default scaling matrix are equal to 16.
Dependencies are inferred from the claim text.