Adaptive DMVR and BDOF Control for Efficient Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in applying bi-directional motion predictors based motion refinement due to lack of adaptive control mechanisms, leading to suboptimal compression and decoding performance.
Innovation Solution
Implement adaptive usage of decoder side motion vector refinement (DMVR) and bi-directional optical flow (BDOF) by applying bi-directional motion predictors based motion refinement only when specific conditions are met, such as threshold comparisons of temporal identification, quantization parameters, and content differences, to optimize video encoding and decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bi-directional motion predictors based motion refinement (DMVR and BDOF) is applied to all current blocks, then motion compensation accuracy is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent changes the parameter of motion refinement application from universal to conditional by introducing multiple criteria (temporal identification difference, quantization parameter difference, content difference measure, scene change detection). These parameter-based filters dynamically determine whether DMVR and BDOF should be applied to each current block, thereby reducing computational complexity while maintaining accuracy where needed
Solution Approach 2:
The patent applies motion refinement selectively to specific current blocks that meet predefined conditions rather than uniformly to all blocks. By evaluating temporal identification, quantization parameters, content differences, and scene changes on a block-by-block basis, the system provides high-quality motion compensation only where necessary, optimizing the trade-off between accuracy and complexity
2Productivity
If adaptive control mechanisms are introduced to selectively apply motion refinement, then computational resources are reduced, but system complexity and processing overhead increase
Solution Approach 1:
The patent performs preliminary evaluations of temporal identification, quantization parameters, content differences, and scene changes before applying motion refinement. By pre-assessing whether current blocks meet the predefined conditions, the system avoids unnecessary computational operations and reduces overall processing overhead despite the added decision-making steps
Solution Approach 2:
The system uses inherently available data (temporal identification, quantization parameters, reference block information) to make self-determination decisions about whether motion refinement should be applied. This self-service approach eliminates the need for external control mechanisms or additional complex decision-making structures
3Reliability
If motion refinement is applied without selective criteria, then encoding consistency is maintained, but compression performance becomes suboptimal
Solution Approach 1:
The patent introduces multiple parameters (temporal identification difference, quantization parameter difference, content difference measure, scene change detection) that dynamically control motion refinement application. This parameter-based approach maintains encoding consistency through systematic decision-making while preventing unnecessary refinement operations that would waste computational resources and degrade compression performance
Data Source
AI summary
Some aspects of the disclosure provide an apparatus for video decoding. The apparatus includes processing circuitry configured to receive a coded video bitstream comprising coded information of one or more pictures, determine, from the coded information, that a current block in a current picture satisfies qualification conditions for a bi-directional motion predictors based motion refinement, start an application of the bi-directional motion predictors based motion refinement on at least a portion of the current block, obtain specific information that is used during the application of the bi-directional motion predictors based motion refinement, and determine, whether to continue the application of the bi-directional motion predictors based motion refinement according to the specific information.