Transform Kernel Derivation for Inter-Coded Blocks with Intra Modes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently combining intra and inter prediction modes for video blocks, leading to suboptimal compression efficiency and quality.
Innovation Solution
The method involves deriving transform kernels based on intra prediction mode information for inter coded blocks, utilizing techniques such as combined inter and intra prediction (CIIP) and geometric partition mode (GPM) to enhance compression by combining intra and inter prediction, and applying transform kernels like DCT and DST for primary and secondary transforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If inter prediction mode is used for current block, then temporal redundancy is exploited for compression, but transform kernel selection becomes suboptimal reducing compression efficiency
Solution Approach 1:
The patent changes the parameter of transform kernel selection by deriving kernels based on intra prediction mode information even for inter-coded blocks. This allows the system to adapt the transform characteristics to the actual content properties (intra prediction mode) rather than being fixed by the coding mode (inter prediction), thereby improving compression efficiency.
Solution Approach 2:
The patent introduces dynamic transform kernel selection where the transform kernels are not fixed but are derived dynamically based on the intra prediction mode information of each block. This dynamic adaptation allows optimal transform kernels to be chosen for each block regardless of whether it is inter-coded or intra-coded, maximizing compression efficiency.
2Productivity
If fixed transform kernels are used for all blocks, then processing is simplified, but compression performance deteriorates due to lack of optimization
Solution Approach 1:
The patent changes the approach from fixed transform kernels to dynamically derived transform kernels based on intra prediction mode information. Although this increases complexity, it significantly improves compression performance by optimizing the transform kernels for each block's actual characteristics rather than using a one-size-fits-all approach.
3Productivity
If intra prediction mode information is used for transform kernel derivation in inter-coded blocks, then compression efficiency improves, but processing complexity increases
Solution Approach 1:
The patent changes the transform kernel derivation process to utilize intra prediction mode information as a guiding parameter. This allows the system to leverage existing intra prediction mode data (already available in the decoding process) to select appropriate transform kernels, improving compression efficiency without requiring entirely new complex processing mechanisms.
Solution Approach 2:
The patent enables the system to use its own existing intra prediction mode information to guide transform kernel selection. Rather than requiring external complex analysis, the system leverages the intra prediction mode data that is already part of the standard decoding process, allowing self-optimization of transform kernels based on available information.
Data Source
AI summary
An aspect of the disclosure provides a method of video decoding. For example, a coded video bitstream is received. The coded video bitstream includes coded information of a plurality of pictures. Based on the coded information, a current block in a current picture is determined to be coded using an inter prediction mode, prediction samples of the current block in the inter prediction mode are generated at least partially based on a reference block in a reference picture that is different from the current picture. Also, intra prediction mode associated with the current block in the current picture is obtained. One or more transform kernels are determined based on the intra prediction mode associated with the current block. The current block is reconstructed based on the one or more transform kernels.