Defect Detection Model Training With Automatic Parameter Configuration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing deep learning algorithm models for defect detection in screen production require manual parameter adjustments, leading to resource wastage and potential human errors.
Innovation Solution
A method for detecting defects using a neural network model that automatically configures training parameters based on feature information, including learning rate descent strategy, total training rounds, and test strategy, reducing the need for human intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual adjustment of training parameters is used, then flexibility in model training is maintained, but human resources are wasted and human errors occur
Solution Approach 1:
The system automatically determines training parameters by analyzing feature information from sample data sets, enabling the model to self-configure without human intervention. This eliminates manual parameter adjustment while maintaining optimal training conditions, resolving the contradiction between reliability and ease of operation.
Solution Approach 2:
The patent dynamically adjusts training parameters based on extracted feature information from sample data, transforming static manual parameter setting into dynamic automated parameter optimization. This resolves the technical contradiction by automatically optimizing parameters to improve defect detection accuracy while eliminating manual intervention.
2Productivity
If automated training parameter configuration is implemented, then human resources are saved and errors are minimized, but system complexity increases
Solution Approach 1:
The system segments the parameter configuration process into distinct automated modules: feature information extraction from sample data sets, analysis of extracted features, and automatic determination of training parameters based on feature analysis. This modular segmentation manages system complexity while achieving automated high-efficiency training.
Solution Approach 2:
Feature information serves as an intermediary between the sample data sets and training parameters. The system extracts feature information from data sets, analyzes these features, and uses them to automatically determine appropriate training parameters. This intermediary mechanism automates the process while managing complexity through structured information transformation.
3Measurement precision
If feature information from sample data sets is used to determine training parameters, then defect detection accuracy is improved, but data processing time increases
Solution Approach 1:
The system performs preliminary extraction and analysis of feature information from sample data sets before model training begins. By pre-processing and identifying key features in advance, the system enables automated parameter determination that improves defect detection accuracy without significantly increasing overall processing time during the training phase.
Data Source
AI summary
A method and device for detecting a defect and method for training a model are provided. The method for detecting the defect includes: acquiring a sample data set and identifying feature information of the sample data set; acquiring an initial model; configuring a training parameter based on the feature information; obtaining a target model by training, according to the training parameter, the initial model with the sample data set; and obtaining defect information of a product by inputting real data of the product into the target model. The training parameter includes at least one of a learning rate descent strategy, a total number of training rounds and a test strategy, the learning rate descent strategy includes a number of learning rate descents and a round number when a learning rate descends, and the test strategy includes a number of tests and a round number when testing.