Multi-Partition Chroma Prediction From Luma in Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently compressing video data while maintaining video quality, particularly in lossy compression scenarios where color components like chroma samples are not fully retained during decoding.
Innovation Solution
Implementing a cross-component prediction (CCP) mode that uses a multi-tap model to determine chroma samples based on luma samples, incorporating nonlinear terms and offsets, and applying model parameters derived from reference area samples to reconstruct chroma blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If lossy compression is applied to reduce data volume, then bandwidth and storage requirements are reduced, but chroma sample quality deteriorates
Solution Approach 1:
The patent applies parameter changes by using a multi-tap cross-component prediction model with adjustable weights and coefficients. The model transforms luma samples through a series of mathematical operations (including nonlinear terms and offsets) to predict chroma samples, allowing flexible adjustment of prediction parameters to optimize both compression ratio and chroma quality based on application requirements
Solution Approach 2:
The patent introduces luma samples as an intermediary to predict chroma samples. Instead of directly encoding chroma data, the system uses the already-encoded luma component as a mediator to derive chroma values through the cross-component prediction model, reducing chroma data volume while maintaining quality through the intermediary prediction process
2Device complexity
If a simple prediction model is used for chroma samples, then decoding complexity is reduced, but prediction accuracy deteriorates
Solution Approach 1:
The patent segments the prediction model into multiple distinct taps or stages, each handling a specific aspect of the prediction process. The multi-tap structure divides the complex prediction task into manageable segments (linear terms, nonlinear terms, offsets) that can be processed sequentially, balancing computational complexity with prediction accuracy
Solution Approach 2:
The patent employs dynamic adaptation by allowing the prediction model parameters to vary across different blocks or regions. The model can adaptively select different tap configurations and parameter sets based on local characteristics, enabling the system to optimize between complexity and accuracy dynamically rather than using a fixed simple or complex model everywhere
3Manufacturing precision
If cross-component prediction is applied to all blocks, then overall video quality is improved, but computational overhead increases
Solution Approach 1:
The patent applies local quality by enabling cross-component prediction selectively for specific blocks or regions where it provides the most benefit. Instead of uniformly applying the prediction model to all blocks, the system identifies regions with specific characteristics (such as high chroma-luma correlation) and applies the prediction technique locally, improving overall quality while reducing unnecessary computational overhead in regions where it would provide minimal benefit
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 a current image frame and a first syntax element for a multi-partition prediction mode. Based on the first syntax element, the computing system determines that the multi-partition prediction mode is enabled to reconstruct the current coding block based on a plurality of partitions. Each of the plurality of partitions corresponds to a set of respective model parameters that is applied to reconstruct a respective chroma sample based on a set of respective luma samples. The computing system determines that a first chroma sample is located in a first partition, combines a set of first luma samples using a set of first model parameters to generate the first chroma sample, and reconstructs the current image frame including the first chroma sample of the current coding block.