Video Decoding Transform Skip for Inter Prediction Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies, particularly for screen content coding (SCC), face limitations in applying the Implicit Selection of Transform Skip mode (ISTS) to inter prediction modes, leading to restricted applicability and increased decoding time.
Innovation Solution
Extend the application of ISTS to inter prediction modes by determining specific conditions such as block size and prediction mode, allowing for inverse transform skip or DCT2 inverse transform processes, and incorporating rate distortion optimization (RDO) to enhance flexibility and controllability of transform skip modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If ISTS is applied only to intra prediction mode and IBC mode, then coding performance is improved for those modes, but applicability is restricted and decoding time increases for other modes
Solution Approach 1:
The patent extends the ISTS mechanism to work across multiple prediction modes (intra prediction, IBC, and inter prediction) by creating a universal transform skip determination framework. The decoding device applies the same ISTS logic regardless of which prediction mode is used, making the system multi-functional and broadly applicable while maintaining coding performance benefits across all modes.
2Measurement precision
If conventional transform operation is performed, then transform accuracy is maintained, but decoding time increases and implementation cost rises
Solution Approach 1:
The patent implements a dynamic transform skip determination mechanism where the decision to perform transform or transform skip is made adaptively based on block characteristics (parity of non-zero transform coefficients) and prediction mode. This dynamic approach allows the system to switch between full transform and simplified transform skip operations, reducing decoding time when transform skip is appropriate while maintaining transform accuracy when needed.
3Productivity
If transform skip process is performed, then decoding time is reduced and implementation cost decreases, but coding performance may deteriorate without proper optimization
Solution Approach 1:
The patent incorporates feedback mechanisms through rate distortion optimization (RDO) that evaluate the effectiveness of transform skip versus full transform based on coding performance metrics. The system uses this feedback to make informed decisions about when to apply transform skip, ensuring that decoding speed improvements do not come at the cost of unacceptable coding performance degradation.
4Adaptability or versatility
If ISTS is extended to inter prediction modes with additional conditions, then flexibility and controllability are enhanced, but device complexity increases
Solution Approach 1:
The patent applies local quality principles by imposing specific conditions (block size less than 64x64, luma block type, specific prediction modes) that are locally appropriate for inter prediction modes. Rather than applying a blanket ISTS rule to all blocks, the system selectively applies transform skip based on local block characteristics, enhancing flexibility and controllability while avoiding unnecessary complexity for blocks that don't meet the criteria.
Data Source
AI summary
Coding and decoding methods and apparatuses are provided. A decoding method includes: determining a prediction mode of a current block; when it is determined that the prediction mode of the current block satisfies one of the following conditions, determining the current block satisfies transform skip mode conditions: the prediction mode of the current block is an inter prediction mode; and the prediction mode of the current block is a direct mode.