Implicit Mode-Dependent Transform Selection for AV1 Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The computational complexity and bitrate overhead associated with selecting hybrid transform kernels in AV1 video coding formats for each residual coding block incur additional processing demands, which can be mitigated by using intra mode-dependent and residual block size-dependent LGT's & KLT's in an implicit manner to improve coding performance.
Innovation Solution
Implementing an implicit method to select hybrid transform kernels based on intra prediction mode and block size, reducing the need for additional signaling and computational complexity in AV1 video coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If hybrid transform kernels are selected for each residual coding block in AV1 video coding, then coding performance is improved, but computational complexity and bitrate overhead increase
Solution Approach 1:
The patent changes the selection parameters from block-level individual selection to mode-level and size-level selection. By using intra prediction mode and residual block size as the basis for selecting hybrid transform kernels, the system reduces the number of independent selection decisions from potentially thousands of blocks to a manageable set of mode-size combinations, thereby reducing computational complexity while maintaining coding performance benefits
Solution Approach 2:
The patent makes the implicit selection method universal across all residual coding blocks by using intra prediction mode and block size as the selection criteria. This universal approach allows the same selection logic to apply to all blocks without requiring block-specific signaling, reducing both computational complexity and bitrate overhead while maintaining improved coding performance
2Manufacturing precision
If hybrid transform kernels are selected for each residual coding block in AV1 video coding, then coding performance is improved, but bitrate overhead increases
Solution Approach 1:
The patent extracts the transform kernel selection information from the block-level syntax and relocates it to the mode and size parameters. By taking out the explicit selection signaling from each block's data stream and replacing it with implicit selection based on already-present intra prediction mode and block size information, the patent eliminates additional bitrate overhead while maintaining coding performance improvements
Solution Approach 2:
The patent enables the intra prediction mode and block size parameters to serve dual purposes: their original function for prediction and classification, plus the additional function of implicitly selecting hybrid transform kernels. This self-service approach allows the existing parameters to carry additional meaning without requiring extra signaling bits, thus improving coding performance without increasing bitrate overhead
Data Source
AI summary
A method, computer program, and computer system is provided for coding video data. Video data is received, and a set of hybrid transform kernels corresponding to the video data is identified. A subset of hybrid transform kernels is selected, either explicitly or implicitly, from among the set of hybrid transform kernels. The video data is decoded based on the selected subset of hybrid transform kernels.