Video Transform Coefficient Coding with Adaptive CABAC Contexts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional context selection for Context-adaptive binary arithmetic coding (CABAC) of coded block flags (CBF) in Versatile Video Coding (VVC) is inefficient, as the probability distribution of CBF values is not dependent on the transform depth or block type, leading to increased implementation cost and potential degradation in probability adaptation.
Innovation Solution
Implementing a simplified context model for CABAC by using a first context model for Cr chroma transform units, a second context model for Cb chroma transform units, and two context modes for luma transform units, with context selection based on the value of the Cb transform unit for the current coding unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional context selection for CABAC is used, then implementation cost is reduced, but coding efficiency and probability adaptation deteriorate
Solution Approach 1:
The patent applies different context models to different chroma components (Cb and Cr) and different transform unit types (luma and chroma). This local differentiation allows the system to adapt to the specific statistical characteristics of each component, improving probability adaptation and coding efficiency without requiring a completely complex new system.
Solution Approach 2:
The context selection process is segmented into distinct categories: luma transform units use one context mode, Cb chroma transform units use another context mode, and Cr chroma transform units use a third context mode. This segmentation simplifies the overall decision logic while maintaining adaptability to different data characteristics.
2Productivity
If context models are not adapted to transform depth or block type, then implementation is simpler, but probability adaptation and coding efficiency worsen
Solution Approach 1:
The patent makes the context model selection dependent on local characteristics such as transform unit type (luma vs. chroma) and chroma component (Cb vs. Cr). This local adaptation allows the probability distribution to match the actual data characteristics better, improving coding efficiency without requiring complex adaptability mechanisms.
Data Source
AI summary
Methods and devices for video encoding are provided. In the method, the device divides a current picture into one or more coding units; determines, for a current luma transform block in a current coding unit of the one or more coding units, whether a first variable indicating an intra subpartition (ISP) split type equals ISP_NO_SPLIT. In response to determining the first variable equals ISP_NO_SPLIT, the device determines a luma block flag context index for a luma block flag of the current luma transform block as being equal to zero or one without determining whether the current luma transform block is smaller than the current coding unit. In response to determining the first variable does not equal ISP_NO_SPLIT, the device determines the luma block flag context index based on a second variable of a previous luma transform unit in the current coding unit.