Method, apparatus, and medium for video processing

Patent metadata

Field	Value
Publication number	US-2026101057-A1
Application number	US-202519406615-A
Country	US
Kind code	A1
Filing date	Dec 2, 2025
Priority date	Jun 2, 2023
Publication date	Apr 9, 2026
Grant date	—

Abstract

Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. In the method, for a conversion between a current video unit of a video and a bitstream of the video, a model parameter inheritance for a coding tool is performed, the coding tool comprising at least one of: a cross-component model based residual coding (CCRM) tool, an intra template matching prediction (intraTMP) filter tool, or an intra block copy (IBC) filter tool, the model parameter inheritance comprising at least one of: a history based model parameter inheritance, a temporal based model parameter inheritance, or a spatial based model parameter inheritance. The conversion is performed based on the model parameter inheritance.

First claim

What is claimed is:

1 . A method for video processing, comprising:
performing, for a conversion between a current video unit of a video and a bitstream of the video, a model parameter inheritance for a coding tool, the coding tool comprising at least one of: a cross-component model based residual coding (CCRM) tool, an intra template matching prediction (intraTMP) filter tool, or an intra block copy (IBC) filter tool, the model parameter inheritance comprising at least one of: a history based model parameter inheritance, a temporal based model parameter inheritance, or a spatial based model parameter inheritance; and performing the conversion based on the model parameter inheritance.

2 . The method of claim 1 , wherein a model parameter set derived by a temporally previous coded video unit is stored at an encoder and a decoder for the conversion, the temporally previous coded video unit comprising at least one of: a temporally previous CCRM coded video unit, a temporally previous intraTMP coded video unit, or a temporally previous IBC coded video unit.

3 . The method of claim 2 , wherein the model parameter set derived by the temporally previous coded video unit inside a current coding tree unit (CTU) or a current CTU row is stored, or
wherein the model parameter set derived by the temporally previous coded video unit inside a current slice or a current tile or a current picture is stored.

4 . The method of claim 1 , wherein a history or temporal or spatial model parameter set is reused by the current video unit, wherein coefficients inside the model parameter set are reused by the current video unit.

5 . The method of claim 1 , wherein a history or temporal or spatial model parameter set list is maintained during the conversion.

6 . The method of claim 5 , wherein the model parameter set list comprises at least one history or temporal model parameter set inside a current coding tree unit (CTU) or CTU row, or
wherein the model parameter set list comprises at least one history or temporal model parameter set inside a current slice or a current tile or a current picture, or wherein the model parameter set list follows a first in first out (FIFO) rule, or wherein the model parameter set list is updated with pruning, or wherein the model parameter set list is reset for each coding tree unit (CTU) or CTU row, or wherein the model parameter set list is reset for each slice, each tile or each picture.

7 . The method of claim 1 , wherein a model parameter set of a CCRM or intraTMP filter coded video unit or an IBC filter coded video unit is stored and used for filtering a subsequent CCRM or intra TMP or IBC coded video unit in a current picture,
wherein the model parameter set is stored in a history-based look-up table.

8 . The method of claim 1 , wherein the model parameter inheritance comprises a spatial adjacent based model parameter inheritance,
wherein a model parameter set derived by a spatial adjacent previous CCRM or intraTMP or IBC coded video unit is stored at an encoder and a decoder for the conversion, and/or wherein a spatial adjacent model parameter set is reused by the current video unit, and/or wherein a spatial adjacent model parameter set list is maintained during the conversion, wherein the model parameter set list comprises at least one spatial adjacent model parameter set inside a current coding tree unit (CTU) or CTU row, or wherein the model parameter set list comprises at least one spatial adjacent model parameter set inside a current slice or a current tile or a current picture, or wherein the model parameter set list follows a first in first out (FIFO) rule, or wherein the model parameter set list is updated with pruning.

9 . The method of claim 1 , wherein the model parameter inheritance comprises a spatial non-adjacent based model parameter inheritance,
wherein a model parameter set derived by a spatial non-adjacent previous CCRM or intraTMP or IBC coded video unit is stored at an encoder and a decoder for the conversion, and/or wherein a spatial non-adjacent model parameter set is reused by the current video unit, and/or wherein a spatial non-adjacent model parameter set list is maintained during the conversion, wherein the model parameter set list comprises at least one spatial non-adjacent model parameter set inside a current coding tree unit (CTU) or CTU row, or wherein the model parameter set list comprises at least one spatial non-adjacent model parameter set inside a current slice or a current tile or a current picture, or wherein the model parameter set list follows a first in first out (FIFO) rule, or wherein the model parameter set list is updated with pruning.

10 . The method of claim 1 , wherein spatial parameters are inherited from at least one of: an adjacent neighboring block, or a non-adjacent neighboring block,
wherein the adjacent neighboring block comprises a block used to inherit motion information in merge mode, or wherein the non-adjacent neighboring block comprises a block used to inherit motion information in merge mode.

11 . The method of claim 1 , further comprising:
determining a model candidate list for the current video unit in a CCRM or intraTMP or IBC model parameter inheritance mode, wherein at least one spatial candidate is inserted in the model candidate list prior to other types of candidates, or wherein candidates in the model candidate list follow an order of spatial adjacent candidates, spatial non-adjacent candidates, history-based candidates, temporal candidates, default candidates, and/or wherein the method further comprises one of:
applying a template cost based reordering process to the model candidate list during a generation of the model candidate list, or
applying a template cost based reordering process to the model candidate list after a construction of the model candidate list.

12 . The method of claim 1 , wherein a first syntax element (SE) is included in the bitstream to indicate whether an inheritance mode is applied,
wherein the first syntax element comprises a flag, or wherein the first syntax element is coded with at least one context model.

13 . The method of claim 1 , wherein a second syntax element is included in the bitstream to indicate which candidate in a candidate list is used,
wherein the second syntax element is binarized as a truncated-unary code or fixed-length code or exponential Golomb code, or wherein a bin of the second syntax element is coded with at least one context model.

14 . The method of claim 1 , wherein the method is applied to a plurality of color components together,
wherein the plurality of color components comprises a Cb component and a Cr component, wherein the plurality of color components shares a same candidate list, or wherein the plurality of color components shares at least one same syntax element.

15 . The method of claim 1 , wherein the method is applied to a plurality of color components separately,
wherein the plurality of color components comprises a Cb component and a Cr component, wherein the plurality of color components has different candidate lists, or wherein the plurality of color components has at least one same syntax element.

16 . The method of claim 1 , wherein a plurality of categories of model parameter sets are jointly inserted into an inheritance list,
wherein the inheritance list is updated following a first in first out rule or a further rule, or wherein the inheritance list is updated with pruning, or wherein the inheritance list is reset for each coding tree unit

Assignees

• Douyin Vision Co Ltd
• Bytedance Inc

Inventors

• Yin Wenbin
• Deng Zhipin
• Zhang Kai
• Zhang Li

Classifications

• H04N19/70

  characterised by syntax aspects related to video coding, e.g. related to compression standards · CPC title

• H04N19/593

  involving spatial prediction techniques · CPC title

• H04N19/196

  being specially adapted for the computation of encoding parameters, e.g. by averaging previously computed encoding parameters (processing of motion vectors H04N19/513) · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title

• H04N19/105

  Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction · CPC title