Palm Print Image Attention Weighting for Unclear Region Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing palm print image classification methods inaccurately extract features from unclear regions, leading to incorrect classification results due to the feature extraction network's excessive attention on these regions.
Innovation Solution
An attention mechanism network is used to recognize attention degrees across different regions of a palm print image, adding predicted labels based on these degrees and correcting its parameters using reference labels to enhance feature extraction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a feature extraction network directly processes the entire palm print image, then the processing is simple and fast, but the extracted embedding feature is inaccurate due to excessive attention on unclear regions
Solution Approach 1:
The palm print image is divided into multiple regions with different definition levels (clear regions and unclear regions). The attention mechanism network segments the image based on definition quality, allowing different processing strategies for different regions. This segmentation enables the system to focus computational resources on clear regions that provide accurate feature information while reducing the impact of unclear regions.
Solution Approach 2:
Different regions of the palm print image are assigned different attention weights based on their definition quality. Clear regions receive higher attention weights while unclear regions receive lower weights. This local quality differentiation ensures that the embedding feature extraction process prioritizes high-quality information while still incorporating lower-quality information in a controlled manner.
2Reliability
If the feature extraction network focuses on all regions equally, then the processing is uniform and simple, but the classification result is inaccurate due to inaccurate features from unclear regions
Solution Approach 1:
Before the main feature extraction and classification process, the attention mechanism network performs preliminary assessment of each region's definition quality and assigns appropriate attention weights. This preliminary action ensures that the subsequent feature extraction process is guided by pre-computed attention weights, improving classification reliability by preventing unclear regions from unduly influencing the results.
Solution Approach 2:
The attention mechanism network uses feedback from the definition assessment of each region to dynamically adjust attention weights during the feature extraction process. Regions with better definition receive positive feedback and higher attention, while regions with poor definition receive negative feedback and lower attention. This feedback mechanism continuously optimizes the feature extraction process to maximize classification accuracy.
Data Source
AI summary
This application discloses an image processing method performed by a computer device. The method includes: obtaining a first sample image including multiple regions; invoking an attention mechanism network to perform attention degree recognition on the regions, to obtain a first region and a second region; adding a first predicted label to the first region and a second predicted label to the second region based on the respective attention degrees of the first region and the second region, and a definition indicated by the first predicted label being higher than a definition indicated by the second predicted label; obtaining a first reference label of the first region and a second reference label of the second region; and updating the attention mechanism network based on a difference between the first predicted label and the first reference label and a difference between the second predicted label and the second reference label.