Neural Loop Filtering With Chroma Fusion for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital video compression technologies struggle to efficiently reduce bandwidth and traffic pressure with the increasing demand for high-quality internet videos, necessitating improved methods to eliminate spatial and temporal redundancies in video data.
Innovation Solution
A video coding method that utilizes neural network-based loop filtering (NNLF) with chroma fusion modes, adjusting chroma information input orders to optimize coding performance by selecting the mode with the minimum rate-distortion cost, and incorporating a chroma adjustment or fusion mode in the NNLF process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If existing digital video compression standards are used, then video data can be saved, but bandwidth and traffic pressure cannot be sufficiently reduced with increasing video clarity requirements
Solution Approach 1:
The patent applies parameter changes by modifying the chroma information processing parameters in the NNLF filter. Specifically, it adjusts the chroma sampling rate and chroma fusion parameters to optimize the balance between compression ratio and video quality, enabling better compression performance while maintaining acceptable quality levels
Solution Approach 2:
The patent replaces traditional mechanical filtering methods with a neural network-based loop filtering system. The NNLF uses learned patterns from training data to perform more efficient compression, substituting conventional signal processing approaches with intelligent algorithms that achieve better compression ratios without proportionally sacrificing quality
2Manufacturing precision
If NNLF with chroma fusion mode is used, then coding performance is enhanced, but computational complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-training the neural network model with augmented training data that includes various chroma information configurations. This pre-training prepares the model to handle different chroma fusion scenarios efficiently during actual encoding, reducing the computational burden during real-time operation while maintaining high coding performance
Solution Approach 2:
The patent introduces dynamic chroma fusion mode selection that adapts to different video content characteristics. The system dynamically adjusts whether to apply chroma fusion and at what level, rather than using a fixed complex processing pipeline, thereby optimizing the balance between coding performance and computational complexity based on actual content requirements
3Manufacturing precision
If chroma information is adjusted in NNLF, then filtering effect is optimized, but processing time increases
Solution Approach 1:
The patent applies local quality by selectively adjusting chroma information only in specific regions or blocks where it provides the most benefit. Rather than uniformly processing all chroma data through complex adjustments, the system identifies and focuses computational resources on areas where chroma optimization will have the greatest impact on overall filtering effect, thereby reducing total processing time
Data Source
AI summary
A video coding method and a storage medium are provided. In terms of performing NNLF on a reconstructed picture, an encoding end can select an optimal mode from a chroma fusion mode and other modes to perform NNLF and set a corresponding flag, where an NNLF model used in the chroma fusion mode is trained using training data obtained by adjusting chroma information.