Decoder-Side Intra Mode Derivation Using Template Candidate Lists
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently compressing video data while maintaining video quality, particularly in limited bandwidth and memory resources, due to inefficiencies in intra prediction modes.
Innovation Solution
The implementation of template-based intra mode derivation (TIMD) for video decoding, which involves obtaining a flag indicating the application of TIMD to a current video block, selecting a template candidate list based on coded information, and adjusting this list using quantization parameters and block sizes, to enhance intra prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If template-based intra mode derivation (TIMD) is applied to handle variations in quantization parameters and block sizes, then coding efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-defining multiple template candidate lists with different characteristics before the actual decoding process. These candidate lists are prepared in advance based on expected quantization parameter ranges and block size variations, allowing the decoder to quickly select an appropriate list without performing complex computations during decoding, thus improving coding efficiency while controlling computational complexity
Solution Approach 2:
The patent utilizes parameter changes by selecting different template candidate lists based on the actual quantization parameter values and block sizes encountered during decoding. When quantization parameters or block sizes fall within certain ranges, the system switches between different pre-defined candidate lists, adapting to varying conditions without requiring complex real-time computations, thereby resolving the contradiction between coding efficiency and computational complexity
2Manufacturing precision
If multiple template candidate lists are maintained for different quantization parameters and block sizes, then intra prediction accuracy is improved, but memory resources increase
Solution Approach 1:
The patent applies segmentation by dividing the template candidate lists into multiple separate lists, each optimized for specific quantization parameter ranges and block size categories. Instead of maintaining one large comprehensive list, the system segments candidates into multiple smaller specialized lists, reducing the memory footprint while maintaining high prediction accuracy for different video block characteristics through targeted selection
Solution Approach 2:
The patent implements local quality by creating template candidate lists with different local characteristics suited for specific conditions. Each candidate list contains templates optimized for particular quantization parameter ranges or block size types, providing locally optimal prediction quality for specific video content characteristics while using minimal memory resources through conditional selection rather than storing all possibilities
3Loss of information
If decoder-side intra mode derivation is implemented to optimize intra prediction, then bit rate is reduced, but processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-organizing template candidate lists and establishing selection criteria before decoding operations begin. The system prepares multiple candidate lists with different characteristics and defines clear selection rules based on quantization parameters and block sizes, enabling the decoder to quickly determine the appropriate list without extensive processing during actual video decoding, thus reducing bit rate while minimizing processing time overhead
Solution Approach 2:
The patent implements self-service through automated template candidate list selection based on decoded quantization parameter values and block size information. The system uses the already-decoded video data parameters to automatically select the most appropriate template candidate list without requiring additional complex processing or external intervention, achieving bit rate reduction through efficient intra prediction while keeping processing time low through self-driven selection based on available information
Data Source
AI summary
Methods, apparatuses, and non-transitory computer-readable storage mediums are provided for video coding. In one method, a decoder obtains a first flag that indicates whether a template-based intra mode derivation (TIMD) mode is applied to a current video block; and in response to determining that the TIMD mode is applied to the current video block based on the first flag, the decoder obtains a set of candidate list for the TIMD mode, where the set of candidate list includes a set of reference line list or a set of template candidate list.