NN Post-Filter Scaling Ratios for Efficient Video Bitstream Signaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards, such as VVC, lack efficient methods for signaling neural-network post-processing filters (NNPFs) and passing activation parameters, which are crucial for enhancing video quality but not required for decoding, leading to inefficiencies in bandwidth usage and video processing.
Innovation Solution
Implementing a method for determining and signaling horizontal and vertical scaling ratios of neural-network post-filters (NNPFs) in a bitstream during spatial resolution upsampling, along with a processor and memory system to execute these parameters, and generating a bitstream based on these ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural-network post-processing filters (NNPFs) are implemented to enhance video quality, then video quality is improved, but bandwidth usage increases due to lack of efficient signaling methods
Solution Approach 1:
The patent extracts only the essential activation parameters (horizontal scaling ratio and vertical scaling ratio) from the complete NNPF configuration, signaling them separately in the bitstream. This selective extraction allows video quality enhancement while minimizing bandwidth consumption by transmitting only the necessary parameters rather than complete filter configurations.
Solution Approach 2:
The patent changes the parameter representation by using scaling ratios (horizontal and vertical) instead of absolute dimension values. This parameter transformation enables more efficient bandwidth utilization while maintaining the ability to accurately control NNPF activation for video quality enhancement.
2Manufacturing precision
If complete NNPF activation parameters are signaled in the bitstream, then video processing accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the NNPF activation parameters into distinct components (horizontal scaling ratio and vertical scaling ratio) and signals them separately in the bitstream. This segmentation maintains video processing accuracy by preserving all necessary parameter information while reducing bitstream complexity through structured, modular parameter organization.
Solution Approach 2:
The patent introduces dynamic parameter signaling where the horizontal and vertical scaling ratios are adaptively determined and signaled based on specific coding conditions. This dynamic approach allows the system to maintain high video processing accuracy when needed while reducing bitstream complexity by only signaling parameters when they deviate from default values.
3Measurement precision
If spatial resolution upsampling is performed with NNPF, then video quality is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary determination of the horizontal and vertical scaling ratios before NNPF activation. By pre-calculating and signaling these parameters, the system prepares the necessary information in advance, enabling faster NNPF processing during video decoding while maintaining high video quality through accurate scaling ratio application.
Solution Approach 2:
The patent enables the decoding system to self-determine the appropriate NNPF activation parameters by signaling the scaling ratios in the bitstream. This self-service mechanism allows the system to automatically adjust processing parameters based on the signaled information, reducing the need for complex external control and optimizing processing time while maintaining video quality.
Data Source
AI summary
A mechanism for processing video data is disclosed. The mechanism includes determining a horizontal scaling ratio of a neural-network post-filter (NNPF) and a vertical scaling ratio of the NNPF are indicated in a bitstream when a NNPF process employs spatial resolution upsampling. A conversion is performed between a visual media data and a bitstream based on the horizontal scaling ratio and the vertical scaling ratio.