Image Feature Encoding via Channel Correlation Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human vision-based encoding methods for data communication in machine learning applications suffer from low code compression rates, leading to difficulties in ensuring communication quality with increased data volumes, particularly in terms of latency and scale.
Innovation Solution
A method for transmitting image features using a machine learning model to extract feature matrices, determine correlated matrices based on correlation thresholds, and encode representative matrices with maximum feature values and minimum feature values to improve code compression and communication quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If human vision-based encoding methods are used for image features, then the encoding process is simple and intuitive, but the code compression rate is low and communication quality deteriorates with increased data volumes
Solution Approach 1:
The patent changes the encoding parameters from human vision-based to machine learning-based feature extraction. By using deep learning models to extract feature matrices and analyzing correlations between different feature channels, the system optimizes the encoding parameters to achieve higher compression rates while maintaining communication quality, directly resolving the contradiction between encoding simplicity and compression efficiency
Solution Approach 2:
The patent creates a simplified representation by copying only the essential correlated feature information rather than transmitting all original data. By identifying correlated feature matrices and transmitting only the representative ones along with correlation parameters, the system reduces data volume while preserving the necessary information for accurate reconstruction
2Reliability
If all feature matrices are transmitted to ensure complete information, then communication quality is maintained, but data volume increases and compression efficiency decreases
Solution Approach 1:
The patent extracts and transmits only the essential correlated feature information by identifying feature matrices with correlation coefficients above a threshold. This extraction process removes redundant information while preserving the core correlated relationships, thereby reducing data volume without compromising communication quality
Solution Approach 2:
The patent merges correlated feature matrices into a unified representation by transmitting the representative feature matrix along with correlation parameters. This combining approach allows the receiver to reconstruct all original features from the compressed representation, maintaining information completeness while reducing transmission data volume
Data Source
AI summary
The invention relates to an image feature transmission method, device and system, and relates to the technical field of communication. The transmission method includes the following steps: extracting a feature matrix of a to-be-processed image for each channel by using a machine learning model; determining one or more incidence matrix pairs according to a comparison result of the correlation degree between the feature matrixes and a first threshold value; according to the information amount, determining a representation matrix and a represented matrix in the two feature matrixes of each incidence matrix pair; determining a corresponding relation between each representation matrix and each represented matrix; and carrying out quantization processing and coding processing on each representation matrix, the corresponding relation and the maximum characteristic value and the minimum characteristic value in each represented matrix, and then transmitting to a decoding end.