Video Frame Compression Using Dual Neural Networks Against Error Drift
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video frame compression methods relying on neural networks accumulate errors frame by frame due to excessive dependence on the quality of the decompressed reference frame, leading to degraded reconstruction quality.
Innovation Solution
A method that utilizes a combination of multiple neural networks, where one network generates features independently of the reference frame and another network uses the reference frame for feature generation, minimizing data transmission while improving reconstruction quality by reducing error accumulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single neural network is used to compress video frames using reference frames, then compression efficiency is improved, but error accumulation occurs frame by frame
Solution Approach 1:
The patent divides the compression system into two independent neural networks: a first neural network that processes current video frames independently without reference frames, and a second neural network that uses reference frames for compression. This segmentation allows each network to operate autonomously, preventing error accumulation while maintaining compression efficiency through specialized functionality.
2Productivity
If reference frames are used for compression, then compression performance is improved, but dependency on reference frame quality increases
Solution Approach 1:
The system segments compression functionality into two independent neural networks. The first neural network processes current video frames independently, eliminating dependency on reference frame quality. The second neural network handles compression using reference frames for scenarios where reference quality is reliable. This segmentation allows the system to maintain compression performance while reducing unnecessary dependency on reference frame quality.
3Reliability
If multiple neural networks are used to combine advantages, then reconstruction quality is improved, but system complexity increases
Solution Approach 1:
The patent segments the neural network system into two distinct networks with specialized functions: the first neural network for independent frame processing and the second neural network for reference-based compression. This segmentation allows each network to be optimized for its specific task, improving overall reconstruction quality while maintaining manageable system complexity through clear functional separation.
Data Source
AI summary
A video frame compression method includes determining a target neural network from a plurality of neural networks according to a network selection policy; and generating, by using the target neural network, compression information corresponding to a current video frame. If the compression information is obtained by using a first neural network, the compression information includes first compression information of a first feature of the current video frame, and a reference frame of the current video frame is used for a compression process of the first feature of the current video frame. If the compression information is obtained by using a second neural network, the compression information includes second compression information of a second feature of the current video frame, and a reference frame of the current video frame is used for a generation process of the second feature of the current video frame.