Geometric Partition Mode Candidate Ordering with Template Matching Costs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in the reordering of merge indices for geometric partition mode (GPM) in video codecs, particularly in the Versatile Video Coding (VVC) standard, leading to suboptimal compression efficiency due to fixed ordering of merge indices without considering template matching costs.
Innovation Solution
The proposed method involves reordering merge indices for GPM partitions based on template matching (TM) costs, using reconstructed templates above and to the left of the current block to reorder candidate reference blocks in ascending order of TM differences, optimizing the merge index selection process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed ordering of merge indices is used in GPM, then device complexity is reduced, but compression efficiency deteriorates
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed, static ordering of merge indices to a dynamic ordering that adapts based on template matching costs. The merge candidate list is reordered according to TM costs calculated for each partition, allowing the system to optimize compression efficiency adaptively without significantly increasing device complexity, as the reordering follows a systematic cost-based approach rather than requiring complex optimization algorithms.
2Loss of energy
If template matching based reordering is implemented, then compression efficiency is improved, but calculation complexity increases
Solution Approach 1:
The patent applies preliminary action by calculating template matching costs for all merge candidates before finalizing the merge index selection. This pre-calculation of TM costs allows the system to establish an optimized ordering in advance, which then guides the selection process. By performing this preparation work beforehand, the actual encoding process benefits from improved compression efficiency without requiring complex real-time optimization during the main encoding operation.
Solution Approach 2:
The patent applies segmentation by dividing the current block into two partitions and calculating template matching costs separately for each partition. This segmented approach allows the system to handle complex calculations in a modular fashion, where each partition's TM costs are computed independently and then used to reorder the merge candidate lists accordingly. This reduces the overall calculation complexity compared to evaluating all candidates simultaneously for the entire block.
Data Source
AI summary
A current block is partitioned into a first partition and a second partition. Template samples of the current block are split into a first template region adjacent to the first partition and a second template region adjacent to the second partition. A plurality of first candidate reference blocks is determined for the first partition. A plurality of second candidate reference blocks is determined for the second partition. At least one of the plurality of first candidate reference blocks and the plurality of second candidate reference blocks is reordered based on a size of the first template region of the template samples and a size of the second template region of the template samples. The current block is reconstructed based on a received index value and based on the reordered at least one of the plurality of first candidate reference blocks and the plurality of second candidate reference blocks.