Selective Multi-Reference-Line Intra-Prediction for Lower Signaling Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques face inefficiencies due to increased signaling overhead when employing multiple reference lines for intra-prediction, leading to larger file sizes and decreased coding efficiency.
Innovation Solution
Implementing an intra-prediction mode subset that allows access to alternative reference lines only for certain modes, limiting others to a primary reference line, and optimizing reference line indexing based on selection probability to reduce signaling overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple reference lines are employed for intra-prediction, then image quality is improved, but signaling overhead increases
Solution Approach 1:
The patent applies local quality by differentiating between two types of intra-prediction modes: those that can access multiple reference lines and those limited to a single primary reference line. This selective approach allows the system to use multiple reference lines only where beneficial for image quality, while avoiding the signaling overhead penalty for modes where it provides little benefit.
Solution Approach 2:
The patent changes the parameter of reference line accessibility based on prediction mode type. Different intra-prediction modes are assigned different reference line configurations - some modes can reference multiple lines while others are constrained to one line, optimizing the balance between quality and overhead.
2Measurement precision
If multiple reference lines are used for all intra-prediction modes, then prediction accuracy improves, but bitstream size increases
Solution Approach 1:
The patent implements local quality by applying multiple reference lines selectively to specific intra-prediction modes rather than universally to all modes. This ensures that prediction accuracy is improved only where it provides meaningful benefits, while avoiding unnecessary bitstream expansion for modes where multiple reference lines would not significantly help.
Solution Approach 2:
The patent applies partial action by providing multiple reference lines to only some intra-prediction modes rather than all modes. This partial application of the multiple reference line feature achieves sufficient prediction accuracy for the majority of cases while keeping the overall bitstream size manageable.
3Adaptability or versatility
If alternative reference lines are made available to all modes, then coding flexibility improves, but device complexity increases
Solution Approach 1:
The patent applies local quality by granting access to alternative reference lines only to specific intra-prediction modes that benefit from this flexibility. This selective approach provides coding flexibility where needed while avoiding the device complexity overhead of implementing and managing multiple reference lines for all modes.
Solution Approach 2:
The patent segments the set of intra-prediction modes into two groups: those that can access multiple reference lines and those that are limited to a single reference line. This segmentation simplifies the overall system design by avoiding the need to manage multiple reference lines uniformly across all modes, thereby reducing device complexity.
Data Source
AI summary
An encoding or decoding method implemented by a video coding device is provided. In an encoding method, the video coding device determines a value of a reference line index of a coding unit, and determines an intra-prediction mode from a list of most probable modes when the value of the reference line index indicates an additional reference line. The video coding device also encodes coding unit into a bitstream based on the intra-prediction mode, and stores the bitstream in a memory of the video coding device. In example decoding method, the video coding device receives a bitstream, determines a value of a reference line index of a coding unit, and determines an intra-prediction mode from a list of most probable modes when a value of the reference line index indicates an additional reference line. The video coding device also decodes the coding unit based on the intra-prediction mode.