Adaptive DMVR Search Range for Video Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies, such as HEVC and VVC, face limitations in coding efficiency and flexibility due to fixed motion vector refinement methods, which can lead to suboptimal performance across different prediction modes and video unit conversions.
Innovation Solution
The proposed solution involves adaptive decoder side motion vector refinement (ADMVR) methods that dynamically determine search ranges and refine motion candidates based on coding information, allowing for iterative refinement and full-pel searches, as well as reference picture resampling to improve coding efficiency and flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If fixed motion vector refinement methods are used in conventional video coding, then device complexity is reduced, but coding efficiency and adaptability deteriorate
Solution Approach 1:
The patent implements dynamic motion vector refinement by determining search ranges adaptively based on prediction mode and video unit characteristics. The search range is adjusted dynamically rather than using a fixed value, allowing the system to adapt to different prediction modes (inter, intra, skip, merge) and different video unit sizes, thereby improving adaptability while maintaining reasonable complexity through controlled dynamic adjustment.
Solution Approach 2:
The patent changes the parameter of search range in the motion vector refinement process based on coding information associated with the video unit. By modifying the search range parameter adaptively according to prediction mode and video unit properties, the system achieves better coding efficiency and adaptability without requiring complete redesign of the refinement mechanism.
2Productivity
If fixed search range is used for DMVR process, then processing speed is maintained, but coding efficiency deteriorates
Solution Approach 1:
The patent changes the search range parameter of the DMVR process based on coding information such as prediction mode and video unit size. By adaptively adjusting the search range rather than using a fixed value, the system achieves better coding efficiency through more accurate motion vector refinement while controlling the refinement time through intelligent parameter selection.
Solution Approach 2:
The patent applies different search range configurations for different prediction modes and video unit types. By segmenting the refinement process into mode-specific configurations, the system optimizes coding efficiency for each mode while avoiding unnecessary refinement operations, thereby balancing coding efficiency with processing time.
3Reliability
If conventional motion candidate selection is used, then device complexity is low, but coding performance deteriorates
Solution Approach 1:
The patent determines motion candidates based on the specific prediction mode and video unit characteristics. By applying different candidate selection strategies locally for different modes (inter, intra, skip, merge) and video unit sizes, the system achieves better coding performance through mode-optimized candidate selection while keeping overall complexity manageable through localized rather than universal complexity.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, during a conversion between a video unit of a video and a bitstream of the video unit, a search range of a decoder side motion vector refinement (DMVR) process based on coding information associated with the video unit; determining a set of motion candidates based on an initial motion candidate and the search range; and performing the conversion based on the set of motion candidates.