Neural Network Post-Filter Signaling in Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards face challenges in efficiently signaling neural network post-filter parameter information for coded video, which affects video quality and compression efficiency.
Innovation Solution
The proposed solution involves signaling neural network post-filter characteristics through specific syntax elements in the bitstream, including indicating the bit depth of luma and chroma sample values output by the post-processing filter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If neural network post-filter parameters are signaled in the bitstream, then video quality is improved, but bitstream complexity increases
Solution Approach 1:
The patent segments the signaling of neural network post-filter parameters by introducing separate syntax elements for different parameter types (e.g., nn_post_filter_enabled_flag, nn_post_filter_bitdepth_luma, nn_post_filter_bitdepth_chroma). This allows selective signaling of only necessary parameters based on configuration needs, reducing overall bitstream complexity while maintaining video quality improvement where applicable.
Solution Approach 2:
The patent applies local quality by allowing different bit depths to be specified for luma and chroma components separately (nn_post_filter_bitdepth_luma and nn_post_filter_bitdepth_chroma). This enables precise control over post-filter processing quality for different color channels independently, optimizing video quality where needed while avoiding unnecessary complexity for channels that don't require enhanced processing.
2Measurement precision
If bit depth information for luma and chroma sample values is signaled, then post-processing filter control precision is improved, but data transmission overhead increases
Solution Approach 1:
The patent introduces dynamic signaling mechanisms where bit depth information for luma and chroma components can be adaptively specified based on the actual video content and processing requirements. The syntax elements nn_post_filter_bitdepth_luma and nn_post_filter_bitdepth_chroma allow the system to dynamically adjust the precision of post-filter control only when and where needed, rather than using fixed high precision throughout, thus reducing data transmission overhead while maintaining control precision where beneficial.
Data Source
AI summary
A device may be configured to perform filtering based on information included in a neural network post-filter characteristics message. In one example, the neural network post-filter characteristics message includes a syntax element specifying the bit depth of luma sample values output by the post-processing filter and a syntax element specifying the bit depth of chroma sample values output by the post-processing filter.