Affine Motion Prediction Weighting for Lower Video Coding Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently compressing video data while maintaining high quality, particularly with affine motion compensation, which increases processing complexity and delays.
Innovation Solution
Implement a method for video coding that includes two alternative prediction weighting processes: default weighted sample prediction and explicit weighted sample prediction, selectively applied based on motion prediction information, to reduce complexity and processing delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If affine motion compensation is used to improve prediction accuracy, then manufacturing precision is improved, but device complexity increases
Solution Approach 1:
The patent applies different weighting processes to different blocks based on their motion prediction mode. Affine inter predicted blocks use a simplified default weighting process, while other blocks use the explicit weighted sample prediction process. This local differentiation maintains prediction accuracy where needed while reducing complexity for affine blocks.
Solution Approach 2:
The patent segments the weighting process into two distinct paths: default weighted sample prediction for affine blocks and explicit weighted sample prediction for other blocks. This segmentation allows the system to apply appropriate complexity only where necessary, resolving the contradiction between accuracy and complexity.
2Manufacturing precision
If affine motion compensation is used to improve prediction accuracy, then manufacturing precision is improved, but processing time increases
Solution Approach 1:
The patent implements local quality by applying the computationally intensive explicit weighting process only to blocks that need it (non-affine blocks), while using the faster default weighting process for affine blocks. This selective approach maintains overall prediction accuracy while significantly reducing total processing time.
Solution Approach 2:
Instead of applying the full explicit weighting process to all blocks, the patent uses partial action by applying it only to non-affine blocks. This partial application of the complex process achieves sufficient prediction accuracy while avoiding the excessive processing time that would result from applying it universally.
3Manufacturing precision
If explicit weighted sample prediction is applied to all blocks, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
The patent applies explicit weighted sample prediction only to blocks where it is most beneficial (non-affine blocks), while using the simpler default process for affine blocks. This local quality approach ensures high prediction accuracy where the complex process is needed while reducing overall device complexity.
Solution Approach 2:
The patent segments the block processing into two categories with different weighting approaches. This segmentation prevents the unnecessary application of complex explicit weighting to affine blocks, thereby reducing overall device complexity while maintaining accuracy where required.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A method of an inter prediction process for a block, wherein the method comprises two alternative prediction weighting processes: a) default weighted sample prediction process, wherein values of weights are redefinable on the block level; and b) explicit weighted sample prediction process, wherein the block is predicted using an explicit weighting process, selecting either process a) or process b) on the basis of motion prediction information of the block, and obtaining the prediction of the block based on the selecting result.