Cross-Component Prediction Models for Chroma Compression Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding methods struggle to effectively utilize redundancy in video data for efficient compression, particularly in lossy scenarios where distortion tolerance varies by application, leading to suboptimal compression ratios and quality trade-offs.
Innovation Solution
Implement multi-hypothesis cross-component prediction (MH-CCP) using a weighted sum of luma samples to predict chroma samples, with weighing factors determined by a least mean square calculation kernel and a multi-tap model, allowing for improved chroma sample reconstruction in video encoding and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If lossy compression is used to reduce data volume, then bandwidth and storage requirements are reduced, but video quality and fidelity deteriorate
Solution Approach 1:
The patent applies parameter changes by using multiple hypothesis models with different weighting factors to predict chroma samples from luma samples. The encoder selects from multiple prediction models (e.g., different tap configurations in the multi-tap filter) and optimizes parameters like weighting factors and model selection to achieve better compression efficiency while maintaining acceptable quality levels appropriate for different applications.
2Productivity
If higher compression ratios are achieved through lossy compression, then bandwidth requirements are reduced, but distortion between original and reconstructed signals increases
Solution Approach 1:
The patent implements dynamics by enabling adaptive selection among multiple hypothesis models based on local image characteristics. The encoder can dynamically switch between different prediction models (e.g., changing the number of taps in the multi-tap filter, adjusting weighting factors) to optimize compression efficiency for different regions and content types, achieving higher overall compression ratios while maintaining quality where critical.
3Productivity
If conventional prediction methods are used, then encoding complexity is kept low, but compression efficiency and ability to utilize video data redundancy is insufficient
Solution Approach 1:
The patent applies segmentation by dividing the prediction process into multiple hypotheses, each using different models or parameters. Instead of a single prediction model, the system segments the prediction task into multiple candidate models (e.g., different multi-tap configurations), evaluates them, and selects the best one. This segmented approach improves compression efficiency by better exploiting local redundancies while keeping individual model complexities manageable.
Solution Approach 2:
The patent implements universality by designing a multi-hypothesis framework where a single prediction structure can accommodate multiple models through parameter variation. The same basic prediction architecture can use different numbers of taps, different weighting schemes, and different model configurations, making the system multi-functional and adaptable to various content types without requiring entirely separate encoding paths for each scenario.
Data Source
AI summary
The various implementations described herein include methods and systems for coding video. In one aspect, a video bitstream includes a current coding block of an image frame and signals a first syntax element for a multi-hypothesis cross-component prediction (MH-CCP) mode. A computing system determines that the MH-CCP mode is enabled to reconstruct each of a plurality of chroma samples of the current coding block using a respective luma sample and one or more neighboring luma samples. The computing device determines a number (N) of model parameters used in the MH-CCP mode for at least the current coding block, and identifies one or more neighboring luma samples of a first luma sample based on the number (N) of model parameters. A first chroma sample that is collocated with the first luma sample is determined based on the first luma sample and the one or more neighboring luma samples.