Video Bitstream NNPF Usage Signaling and SEI Message Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards, such as VVC and VSEI, include SEI messages for neural-network post-filters that are not essential for decoding but are not efficiently managed, leading to inefficiencies in video processing and bandwidth usage.
Innovation Solution
The method involves signaling indications of usage suitability for neural-network post-filters and removing unnecessary SEI messages in sub-bitstream extraction, optimizing video processing by eliminating redundant information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If SEI messages for neural-network post-filters are included in video bitstreams, then post-processing functionality is provided, but bandwidth usage increases and processing efficiency decreases
Solution Approach 1:
The patent extracts and removes SEI messages related to neural-network post-filters from the video bitstream during encoding. By taking out these optional post-processing instructions, the system reduces the bitstream size and bandwidth requirements while maintaining the core video coding functionality, thus resolving the contradiction between providing post-processing capabilities and minimizing bandwidth usage.
Solution Approach 2:
The patent segments the video coding process into essential decoding components and optional post-processing components. By separating these functions and only including SEI messages when necessary, the system achieves flexible bandwidth usage that adapts to different application requirements, resolving the contradiction between versatility and bandwidth efficiency.
2Adaptability or versatility
If SEI messages for neural-network post-filters are included in video bitstreams, then post-processing functionality is provided, but data transmission volume increases
Solution Approach 1:
The patent extracts and removes SEI messages related to neural-network post-filters from the video bitstream during encoding. By taking out these optional post-processing instructions, the system reduces the bitstream size and bandwidth requirements while maintaining the core video coding functionality, thus resolving the contradiction between providing post-processing capabilities and minimizing bandwidth usage.
3Loss of information
If all SEI messages are processed during decoding, then complete video information is obtained, but processing time increases
Solution Approach 1:
The patent applies preliminary action by removing SEI messages related to neural-network post-filters during the encoding stage before the bitstream is transmitted or stored. This preliminary removal reduces the amount of data that needs to be processed during decoding, thereby reducing processing time while maintaining information completeness for essential video coding functions.
Solution Approach 2:
The patent extracts and removes SEI messages related to neural-network post-filters from the video bitstream during encoding. By taking out these optional post-processing instructions, the system reduces the bitstream size and bandwidth requirements while maintaining the core video coding functionality, thus resolving the contradiction between providing post-processing capabilities and minimizing bandwidth usage.
Data Source
AI summary
A mechanism for processing video data is disclosed. The mechanism includes determining an indication that indicates information related to usage for a neural-network post-filter (NNPF). The indication may indicate that a suitable usage for the NNPF or a video bitstream is unknown. A conversion is performed between a visual media data and a bitstream based on the indication.