Cross-Component Intra Prediction With Chroma Phase Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cross-component intra prediction methods in video coding suffer from high memory footprint and computational complexity, limiting their efficiency and flexibility.
Innovation Solution
Implementing a method that signals phase values between chroma and luma components, allowing for shared or individual phase information, and using convolutional filters independent of chroma format, while reusing prediction model parameters across blocks to reduce computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cross component intra prediction is applied to improve compression efficiency, then prediction accuracy is improved, but memory footprint and computational complexity increase
Solution Approach 1:
The picture is divided into multiple portions with different chroma phases, allowing independent processing of each portion. This segmentation enables the use of simplified prediction methods for each phase while maintaining overall prediction accuracy, thereby reducing computational complexity compared to applying complex prediction to the entire picture.
Solution Approach 2:
The patent changes the parameter of chroma phase by signaling different phase values (e.g., phase 0, phase 1, phase 2) for different portions of the picture. This parameter change allows the system to adapt to different chroma-luma relationships in different regions, improving prediction accuracy while using simpler, phase-specific prediction methods rather than universally complex methods.
2Measurement precision
If cross component intra prediction is applied to improve compression efficiency, then prediction accuracy is improved, but memory footprint increases
Solution Approach 1:
By segmenting the picture into portions with different chroma phases and processing each phase independently, the patent reduces the amount of reference data that needs to be stored in memory. Each phase can use its own simplified reference model rather than requiring all reference data for all phases, thereby reducing memory footprint while maintaining prediction accuracy.
Solution Approach 2:
The signaling of chroma phase parameters allows the system to select appropriate reference models for different portions of the picture. This parameter-based approach enables memory-efficient processing by only storing and processing the necessary reference data for each specific phase, rather than maintaining large memory structures for all possible phase combinations.
3Adaptability or versatility
If chroma phase information is signaled for each chroma component to improve flexibility, then adaptability is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal chroma phase signaling mechanism that can be applied to different chroma components (Cb and Cr) and different picture portions. This multi-functional approach allows the same phase signaling infrastructure to serve multiple purposes and components, improving flexibility while avoiding the need for separate complex signaling systems for each component, thereby controlling device complexity.
Data Source
AI summary
A method for signaling or parsing. for each chroma component of at least one portion of a picture. an information representative of a phase value between samples of the chroma component and samples of a luma component of the picture, the phase value being taken among at least three different phase values for each chroma component.