Decoder-Side Motion Vector Refinement Threshold
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently refining motion vectors at the decoder side, particularly for larger blocks of pixels, which can increase computational workload and processing time without ensuring significant video quality improvement.
Innovation Solution
The proposed solution involves a video codec that performs decoder-side motion vector refinement (DMVR) by identifying initial motion vectors for a current block of pixels and refining them to minimize distortion cost, with the refinement process being conditional on the block size relative to a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If motion vector refinement is performed for all blocks regardless of size, then motion vector precision is improved, but computational workload increases significantly
Solution Approach 1:
The patent changes the parameter of block size as a decision criterion. By comparing block size against a threshold, the system dynamically adjusts whether to perform motion vector refinement, thereby optimizing the balance between precision and computational efficiency based on the specific characteristics of each block.
Solution Approach 2:
Instead of applying motion vector refinement universally to all blocks, the patent applies it partially - only to blocks that meet specific size criteria. This partial action approach avoids unnecessary computational overhead for blocks where refinement would not provide significant quality improvement.
2Manufacturing precision
If motion vector refinement is performed for large blocks, then video quality is improved, but processing time increases
Solution Approach 1:
The patent uses block size as a changing parameter to control the refinement process. By setting a threshold value, the system adapts its processing behavior - performing refinement only when the block size justifies the additional processing time, thus optimizing the video quality to processing time ratio.
3Measurement precision
If motion vector refinement is always performed, then coding accuracy is improved, but computational resources are wasted
Solution Approach 1:
The patent introduces block size as a dynamic parameter that controls resource allocation. By comparing the block size parameter against a threshold, the system intelligently decides whether to allocate computational resources for refinement, avoiding waste on blocks where refinement would not yield significant accuracy improvements.
Solution Approach 2:
The system performs self-assessment by evaluating block size characteristics before committing to refinement operations. This self-service mechanism allows the coding process to automatically allocate resources based on actual needs, improving overall efficiency without external intervention.
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
A video codec that encodes or decodes video sequences using decoder-side motion vector refinement is provided. The video codec identifies a first motion vector and a second motion vector for coding a current block of pixels of a current video frame in the video sequence. The video codec determines whether to perform motion vector refinement for the current block of pixels based on a comparison between a linear dimension of the current block of pixels and a threshold. When motion vector refinement is performed, the video codec refines the first and second motion vectors to minimize a distortion cost and codes the current block of pixels by using the refined first and second motion vectors. When motion vector refinement is not performed, the video codec codes the current block of pixels by using the identified first and second motion vectors.