Video Decoder Weight Determination for Merge Mode
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face challenges in efficiently determining prediction weights for merge modes, particularly in combining different techniques applicable to various motion vector candidates, leading to inconsistent weight inheritance and reduced coding performance.
Innovation Solution
The proposed solution involves determining weights for inherited affine merge candidates, constructed affine merge candidates, or zero motion vectors in video data processing. This is achieved by signaling a weight index in the bitstream or deriving weights based on control points, allowing for flexible weight selection and signaling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If different weight determination techniques are applied to various motion vector candidates, then coding performance is improved, but device complexity increases due to the need to properly combine and select multiple techniques
Solution Approach 1:
The patent applies parameter changes by dynamically selecting different weight determination techniques based on the type of motion vector candidate (inherited affine merge candidate, constructed affine merge candidate, or zero motion vector). The decoder changes the parameter of weight determination method according to the candidate type, allowing optimal performance for each case while maintaining manageable complexity through systematic classification.
2Measurement precision
If explicit weight signaling is implemented, then weight determination accuracy is improved, but loss of information increases due to additional bitstream data requirements
Solution Approach 1:
The patent applies local quality by signaling weight indices selectively based on the type of motion vector candidate. Instead of uniformly signaling weights for all candidates, the system applies explicit weight signaling only where needed (for inherited and constructed affine merge candidates) while using default or derived weights for zero motion vectors. This localized approach improves accuracy where necessary while minimizing bitstream overhead.
3Ease of operation
If weight inheritance is applied to all motion vector candidates, then ease of operation is improved, but manufacturing precision deteriorates due to inconsistent weight determination
Solution Approach 1:
The patent applies segmentation by dividing motion vector candidates into distinct categories (inherited affine merge candidates, constructed affine merge candidates, and zero motion vectors) and applying appropriate weight determination methods to each segment. This segmentation resolves the contradiction by allowing simple weight inheritance for some candidates while applying more precise determination methods for others, achieving both ease of operation and manufacturing precision through structured classification.
Data Source
AI summary
The present disclosure provides systems, devices, and methods for determining prediction weight for sub-block. One exemplary method comprises: partitioning a coding unit into a plurality of sub-blocks; determining a weight of a sub-block of the plurality of sub-blocks based on weights of control points, wherein the control points are neighboring corners of the coding unit; and bi-predicting the sub-block based on the determined weight