Video Decoder Template Selection for Intra-Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding technologies rely on fixed templates for intra-prediction, which may not optimally predict block samples, leading to inefficient bitrate and processing due to the inclusion of unavailable samples and fixed IPM signaling.
Innovation Solution
The method involves constructing a template set from a plurality of sub-templates, such as left, above, right-above, left-below, and left-above sub-templates, based on block dimensions and coding information, to derive an intra-prediction mode (IPM) and determine a final predictor for video blocks, reducing the processing of unavailable samples and improving prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed templates are used for intra-prediction, then the signaling process is simplified, but the prediction accuracy decreases and bitrate efficiency is reduced
Solution Approach 1:
The patent implements dynamic template selection where the template set is constructed adaptively based on block characteristics such as dimension ratios. Different template configurations (e.g., horizontal, vertical, or combined templates) are selected according to the specific block geometry, allowing the system to optimize prediction accuracy for each block type while maintaining manageable signaling complexity through conditional logic.
Solution Approach 2:
The patent changes the parameters of the template construction process by adjusting which sub-templates are included in the final template set based on block dimension parameters. The template selection depends on comparing block width-to-height ratios against thresholds, dynamically modifying the template configuration to match the block shape, thereby improving prediction accuracy without requiring complex signaling.
2Adaptability or versatility
If fixed templates include all sub-templates, then comprehensive coverage is achieved, but processing of unavailable samples increases
Solution Approach 1:
The patent extracts only the necessary sub-templates from the full set of available templates based on block characteristics. For example, for wide blocks (width-to-height ratio above threshold), only horizontal templates are constructed, excluding vertical templates that would require processing unavailable samples. This selective extraction maintains comprehensive coverage where needed while eliminating unnecessary processing.
Solution Approach 2:
The patent applies different template construction strategies to different block types based on their local characteristics. Each block is analyzed individually, and the template set is customized to match its specific geometry and available reference samples. This local adaptation ensures that each block receives the most appropriate template configuration without universally processing all possible templates.
3Measurement precision
If template sets are constructed based on block characteristics, then prediction accuracy is improved, but the processing complexity increases
Solution Approach 1:
The patent segments the template construction process into distinct stages: first analyzing block characteristics (dimension ratios), then selecting appropriate template configurations, and finally constructing the template set. This segmentation allows the complexity to be managed through modular processing steps, where each stage handles a specific aspect of the problem, making the overall process more tractable while maintaining high prediction accuracy.
Solution Approach 2:
The patent uses simple parameter comparisons (block width-to-height ratio against predefined thresholds) to drive template selection, avoiding complex analysis. This parameter-based approach maintains low processing complexity by relying on straightforward geometric calculations rather than sophisticated algorithms, while still achieving adaptive template construction that improves prediction accuracy.
Data Source
AI summary
Example implementations include a method, apparatus and computer-readable medium of video processing, including constructing, during a conversion between a current video block of a video and a bitstream of the video, at least one template set for the current video block from a plurality of sub-templates. The one or more sub-templates may be selected from a plurality of sub-templates including: a left sub-template, an above sub-template, a right-above sub-template, a left-below sub-template, and a left-above sub-template. The implementations further include deriving at least one intra-prediction mode (IPM) based on cost calculations. The implementations include determining, based on the at least one IPM, a final predictor of the current video block. The implementations include performing the conversion based on the final predictor.