HMVP Table Update Control in Video Coding Merge Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding techniques face inefficiencies in reducing redundancy and bandwidth requirements, particularly in inter-picture prediction, where updating the history-based motion vector prediction (HMVP) table for each inter-coded block within a merge sharing region is redundant and computationally costly, especially when blocks share similar motion information.
Innovation Solution
The method involves constructing a shared merge candidate list for a merge sharing region and decoding it without updating the HMVP table with motion information for all inter-coded blocks, except possibly the first or last block in decoding order, to reduce redundant updates and computational costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the HMVP table is updated with motion information for each inter-coded block within a merge sharing region, then the motion prediction accuracy is improved, but the computational complexity and processing time increase significantly
Solution Approach 1:
The current picture is divided into merge sharing regions, where only the first inter-coded block updates the HMVP table and subsequent blocks share the motion information. This segmentation approach reduces redundant updates while maintaining prediction accuracy within each region.
Solution Approach 2:
Multiple inter-coded blocks within a merge sharing region are merged to share the same motion information from the first block. This combining strategy eliminates redundant HMVP table updates while preserving the benefits of motion compensation for all blocks in the region.
2Manufacturing precision
If the HMVP table is updated for every inter-coded block, then the coding precision is improved, but the processing speed decreases
Solution Approach 1:
The picture is segmented into merge sharing regions with a predetermined number of blocks. Within each region, only the first block performs HMVP table updates, while subsequent blocks share the motion information, thereby maintaining coding precision while improving processing speed.
Solution Approach 2:
Instead of updating the HMVP table for every inter-coded block (excessive action), the patent applies partial action by updating only for the first block in each merge sharing region. This partial approach suffices to maintain coding precision while significantly reducing processing overhead.
3Device complexity
If motion information is shared among multiple blocks in a merge sharing region, then the computational overhead is reduced, but the motion information accuracy for individual blocks may deteriorate
Solution Approach 1:
The patent applies local quality by ensuring that within each merge sharing region, the first block's motion information (which is most relevant to that local area) is used for all subsequent blocks in the region. This local approach maintains accuracy where it matters most while reducing overall computational overhead.
Data Source
AI summary
A method of video encoding includes receiving a merge sharing region including a plurality of coding blocks, constructing a shared merge candidate list for the merge sharing region, and encoding a current inter coded coding block in the merge sharing region based on the shared merge candidate list. The method also includes determining whether to update a history-based motion vector prediction (HMVP) table with motion information of the current inter coded coding block based on whether the current inter coded coding block is inter coded with a merge/skip mode. The method further includes updating the HMVP table with the motion information of the current inter coded coding block when the HMVP table is determined to be updated with the motion information of the current inter coded coding block.