Selective Decoder-Side Motion Vector Refinement for Efficient Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of efficiently compressing video data for transmission and storage with minimal loss in image quality, particularly in scenarios with limited network resources, is addressed by improving decoder-side motion vector refinement (DMVR) techniques in video coding.
Innovation Solution
The method involves determining the temporal relationship between current and reference pictures to perform motion vector refinement (DMVR) when specific temporal distances are equal, using initial motion vectors to refine reference blocks, and applying template matching to optimize motion compensation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If video data is compressed to reduce data size for transmission and storage, then network bandwidth requirements and storage needs are reduced, but image quality may deteriorate
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the temporal distance threshold for applying DMVR based on picture type (I, P, or B frame) and temporal distance values. Different threshold criteria are used for different frame types to optimize the balance between compression efficiency and image quality preservation.
Solution Approach 2:
The system dynamically determines whether to apply DMVR based on real-time conditions including picture type, temporal distance, and threshold comparisons. The motion vector refinement is conditionally applied rather than consistently applied, allowing adaptive optimization of compression quality ratio.
2Productivity
If decoder-side motion vector refinement is applied to improve compression efficiency, then coding efficiency and compression ratio are improved, but computational complexity at the decoder increases
Solution Approach 1:
The patent applies partial action by selectively applying DMVR only to specific picture types (I, P, or B frames) and only when temporal distance conditions are met. Not all blocks undergo full DMVR processing, reducing the overall computational burden while maintaining efficiency improvements where beneficial.
Solution Approach 2:
The system performs preliminary determination of temporal distances and threshold comparisons before actually executing the DMVR algorithm. This preliminary filtering avoids unnecessary computational complexity by identifying in advance which blocks qualify for DMVR processing based on their temporal characteristics.
3Measurement precision
If motion vector refinement is applied to all blocks, then prediction accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent applies local quality by treating different blocks differently based on their specific characteristics. Blocks that meet the temporal distance and picture type criteria receive full DMVR processing with higher precision, while other blocks use simplified or alternative processing methods, optimizing the balance between accuracy and speed.
Solution Approach 2:
Instead of applying motion vector refinement uniformly to all blocks, the system applies it partially only to blocks that satisfy specific conditions (temporal distance thresholds, picture types). This selective application reduces processing time while maintaining high accuracy where the conditions are met.
Data Source
AI summary
A method for inter-prediction of a current image block in a current picture of a video is provided. The method includes determining whether a first temporal distance (such as TD0) is equal to a second temporal distance (such as TD1), wherein the first temporal distance is represented in terms of a difference between a picture order count value of the current picture and a picture order count value of a first reference picture; and the second temporal distance is represented in terms of a difference between a picture order count value of a second reference picture and the picture order count value of the current picture; and performing no motion vector refinement (DMVR) procedure when it is determined that the first temporal distance (TD0) is not equal to the second temporal distance (TD1). Thus the DMVR procedure is restricted to only the image block with equal-distance references.