NN Post-Filter Resampling for Variable-Size Image Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to a surge in transmission and storage costs due to higher bit rates, necessitating a high-efficiency image compression technology.
Innovation Solution
An image encoding/decoding method that includes resampling an input picture of a neural-network post-filter (NNPF) based on the size information of the current picture, allowing for efficient encoding/decoding of images with different sizes, and generating a bitstream that can be stored and transmitted effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-resolution and high-quality images are transmitted, then image quality is improved, but transmission cost and storage cost increase due to higher bit rate
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the neural network post-filter activation status based on image quality metrics and transmission requirements. The system modifies filtering parameters and neural network processing intensity to optimize the balance between image quality preservation and bit rate reduction, thereby lowering transmission and storage costs while maintaining acceptable image quality.
2Manufacturing precision
If neural-network post-filter is applied to enhance image quality, then image quality is improved, but processing time and computational complexity increase
Solution Approach 1:
The patent implements dynamics by making the neural-network post-filter activation adaptive rather than static. The system dynamically activates or deactivates the post-filter based on real-time assessment of image quality metrics and transmission conditions. This dynamic adjustment allows the system to apply computational resources only when necessary, reducing processing time while maintaining image quality enhancement where beneficial.
Solution Approach 2:
The system changes processing parameters by adjusting the intensity and application of neural network filtering based on input image characteristics and transmission requirements. By modulating filter strength, processing resolution, and activation timing, the system optimizes the trade-off between image quality improvement and computational processing time.
3Loss of energy
If image compression is increased to reduce bit rate, then transmission cost is reduced, but image quality deteriorates
Solution Approach 1:
The patent implements feedback mechanisms that continuously monitor image quality metrics and transmission parameters. Based on this feedback, the system adjusts compression levels and neural network post-filter application in real-time. The feedback loop enables dynamic optimization of the compression-quality trade-off, allowing the system to achieve lower bit rates while maintaining acceptable image quality through adaptive control.
Solution Approach 2:
The system applies parameter changes by dynamically adjusting compression parameters and neural network processing intensity based on transmission requirements and image characteristics. By modulating these parameters in real-time, the system optimizes the balance between bit rate reduction and image quality preservation, achieving cost-effective transmission without excessive quality loss.
Data Source
AI summary
An image encoding/decoding method and device are provided. The image decoding method according to the present disclosure may comprise the steps of: acquiring size information of a current picture; and resampling an input picture of a neural-network post-filter (NNPF) on the basis of the size information.