Spatial Geometric Partition Intra Prediction for Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems face inefficiencies in intra prediction modes, particularly in deriving and utilizing spatial geometric partition modes for coding blocks, leading to suboptimal compression and decoding performance.
Innovation Solution
A video decoding device derives multiple candidate intra prediction modes based on different templates for a coding block, adds them to a candidate list, and selects a third candidate mode based on the block's partition mode, utilizing spatial geometric partition mode (SGPM) to enhance prediction accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple candidate intra prediction modes are derived and added to the IPM candidate list, then prediction accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the intra prediction mode derivation process into multiple independent template-based derivation paths. Each template (first template, second template, third template) corresponds to a specific geometric partition configuration, allowing the system to systematically generate multiple candidate IPMs without overwhelming complexity. The segmentation enables parallel processing of different prediction candidates while maintaining organized complexity management through structured template application.
Solution Approach 2:
The patent applies preliminary action by pre-defining multiple templates before the actual prediction process. These templates are prepared in advance with specific geometric configurations and partition modes, allowing the decoding device to quickly generate candidate IPMs without complex real-time calculations. The templates serve as pre-computed reference structures that simplify the prediction mode derivation process while ensuring comprehensive coverage of possible prediction scenarios.
2Productivity
If spatial geometric partition mode is used to derive multiple candidate IPMs, then compression efficiency is improved, but processing time increases
Solution Approach 1:
The patent utilizes parameter changes by dynamically adjusting the template selection and IPM derivation process based on the specific coding block characteristics. The system changes parameters such as template type, partition mode, and candidate IPM generation intensity according to the block's spatial geometry and content characteristics. This adaptive parameter adjustment allows the system to optimize compression efficiency for each block while controlling processing time by avoiding unnecessary derivations for blocks that can be efficiently predicted with fewer candidates.
Solution Approach 2:
The patent applies partial action by deriving and evaluating only the necessary number of candidate IPMs for each coding block. Rather than exhaustively testing all possible prediction modes, the system generates a limited set of promising candidates based on template matching and geometric partition analysis. This partial action approach maintains high compression efficiency by focusing computational resources on the most likely correct prediction modes while avoiding the time cost of evaluating all possible options.
3Measurement precision
If multiple templates are used for IPM derivation, then prediction accuracy is improved, but algorithm complexity increases
Solution Approach 1:
The patent implements universality by designing a unified template-based framework that can generate multiple candidate IPMs across different geometric partition configurations. The same fundamental template derivation mechanism is applied universally to various block types and partition modes, eliminating the need for separate complex algorithms for each scenario. The multi-functional template system handles diverse prediction needs through a single coherent approach, reducing overall algorithmic complexity while maintaining high prediction accuracy across different video content types.
Data Source
AI summary
Systems, methods, and instrumentalities are disclosed associated with spatial geometric partition mode (SGPM). In examples, a device (e.g., a video encoder and/or a video decoder) may derive a first candidate intra prediction mode (IPM) based on a first template of a coding block. The device may derive a second candidate IPM based on a second template of the coding block. The device may process the coding block based on the first candidate IPM and the second candidate IPM.