Adaptive Extrapolation Filter Intra Prediction by Block Size
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently utilizing spatial redundancy for intra prediction, particularly in determining optimal template and filter types for extrapolation filter-based intra prediction (EIP) modes, leading to suboptimal compression efficiency.
Innovation Solution
A method and apparatus for video encoding/decoding that selects template and filter types based on size information and threshold values for extrapolation filter-based intra prediction (EIP) modes, allowing for dynamic determination of predicted values using available template and filter types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional intra prediction methods are used, then device complexity is reduced, but compression efficiency deteriorates due to suboptimal utilization of spatial redundancy
Solution Approach 1:
The patent applies parameter changes by dynamically selecting template types (TPL0, TPL1, TPL2) and filter types (F0, F1, F2) based on block size parameters (width w, height h, area w×h, ratio w/h). Different parameter ranges trigger different template-filter combinations, optimizing prediction accuracy for various block sizes while maintaining manageable complexity through systematic parameter mapping.
Solution Approach 2:
The invention implements dynamics by making the template type and filter type selection adaptive rather than static. The available template types and filter types are determined dynamically based on the current block's size information and predefined threshold values, allowing the prediction mechanism to adapt to different spatial redundancy patterns in various block configurations.
2Measurement precision
If fixed template and filter types are used for EIP modes, then device complexity is reduced, but prediction accuracy deteriorates
Solution Approach 1:
The patent improves prediction accuracy by changing parameters (template type and filter type) according to block size characteristics. By establishing multiple template types (TPL0 with both above and left samples, TPL1 with only above samples, TPL2 with only left samples) and multiple filter types (F0 square, F1 horizontal, F2 vertical), the system can select the most appropriate combination for each block, thereby improving measurement precision of intra prediction.
Solution Approach 2:
The invention applies local quality by making the template and filter selection specific to each block's local characteristics (size, shape, position). Different regions of the video signal with different block size properties receive customized template-filter combinations, rather than applying a uniform approach, thus improving local prediction accuracy where it matters most.
3Measurement precision
If multiple template types and filter types are available for all block sizes, then prediction accuracy is improved, but loss of information increases due to additional syntax elements in bitstream
Solution Approach 1:
The patent reduces bitstream overhead by changing the parameter set (available template types and filter types) based on block size conditions. Instead of always signaling all possible template and filter combinations, the system selectively enables subsets based on predefined threshold values, thereby reducing the number of syntax elements that need to be transmitted while maintaining prediction accuracy for relevant block sizes.
Solution Approach 2:
The invention applies partial action by providing full template and filter type availability only when needed (for certain block size ranges), and limiting the available set for other cases. This selective provision of resources reduces unnecessary information transmission while ensuring that when full capability is required, it is available.
Data Source
AI summary
Aspects of the disclosure includes methods and apparatuses for video decoding and video encoding and a method of processing visual media data. The apparatus for video decoding includes processing circuitry configured to: receive coded information indicating that a current block in a current picture is predicted according to an extrapolation filter-based intra prediction (EIP) mode; determine at least one of (i) one or more available template types from a set of template types and (ii) one or more available filter types from a set of filter types used in the EIP mode based on size information of the current block and at least one threshold value; determine a predicted value of a current sample in the current block according to the EIP mode with the at least one of (i) the one or more available template types and (ii) the one or more available filter types.