Neural-Network Post-Filter Chroma Format Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards face challenges in efficiently processing and transmitting digital video data, particularly in managing neural-network post-filters and chroma format conversions within video bitstreams.
Innovation Solution
The proposed method involves determining the value of the neural-network post-filter characteristics (NNPFC) output chroma component block flag in a Supplemental Enhancement Information (SEI) message, specifically setting it to 1 when the chroma format identification code is 2, indicating a 4:2:2 format, and performing conversions between visual media data and bitstreams based on this flag.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If chroma format conversion is performed in 4:2:2 format with neural-network post-filter, then video quality is improved, but bandwidth consumption increases
Solution Approach 1:
The patent changes the chroma format parameter from 4:2:0 to 4:2:2 to improve video quality. By modifying the chroma subsampling ratio parameter, the system achieves better color fidelity and image quality while managing the associated bandwidth increase through efficient neural-network post-filter processing.
2Measurement precision
If neural-network post-filter is applied to enhance video processing, then video quality improves, but processing complexity increases
Solution Approach 1:
The patent applies preliminary chroma format conversion to 4:2:2 before neural-network post-filter processing. This preliminary action optimizes the input data structure for the neural network, enabling more efficient processing and reducing the computational complexity that would otherwise be required to achieve the same quality improvement.
3Measurement precision
If chroma format is set to 4:2:2 for better color representation, then color accuracy improves, but data transmission volume increases
Solution Approach 1:
The patent modifies the chroma format parameter to 4:2:2 to achieve superior color accuracy and representation. This parameter change increases the chroma component resolution from 4:2:0, providing better color fidelity while the system manages the increased data volume through optimized neural-network post-filter processing.
Data Source
AI summary
A mechanism for processing video data is disclosed. The mechanism includes determining a value of a neural-network post-filter characteristics (NNPFC) output chroma component block flag (nnpfc_out_sub_c_flag) in a NNPFC Supplemental Enhancement Information (SEI) message, wherein when a chroma format identification code (ChromaFormatIdc) is equal to two, which indicates that a colour format is a 4:2:2 format, and a NNPFC output chroma component block flag (nnpfc_out_sub_c_flag) is present, the value of nnpfc_out_sub_c_flag shall be equal to one. A conversion is performed between a visual media data and a bitstream based on the nnpfc_out_sub_c_flag.