Steel surface defect detection is a key part of the production process in the steel industry.The traditional manual inspection methods are inefficient and costly.With the rapid development of deep learning technology,...Steel surface defect detection is a key part of the production process in the steel industry.The traditional manual inspection methods are inefficient and costly.With the rapid development of deep learning technology,automatic detection and segmentation of steel surface defects based on deep neural networks has received widespread attention and demonstrated good performance in several real-world scenarios.However,challenges remain due to the obvious inter-class similarity and intra-class variation problems in steel defect images,impacting accuracy and robustness.This paper systematically summarizes the representative researches on solving the inter-class similarity and intra-class variation problems in recent years,and focuses on analyzing the innovations of different methods in network structure design.In addition,this paper also discusses the shortcomings of the current research,and proposes a new idea of fusion of mainstream modeling method.By employing a subcomparison module to enhance feature similarity within classes and differentiate across classes,followed by pyramid feature fusion to optimize computational efficiency,this study aims to advance high-precision intelligent recognition of steel surface defects.This study reveals that the approach not only addresses existing challenges but also provides a foundation for future advancements in steel defect detection technologies.展开更多
Spectrum-based fault localization (SBFL) generates a ranked list of suspicious elements by using the program execution spectrum, but the excessive number of elements ranked in parallel results in low localization accu...Spectrum-based fault localization (SBFL) generates a ranked list of suspicious elements by using the program execution spectrum, but the excessive number of elements ranked in parallel results in low localization accuracy. Most researchers consider intra-class dependencies to improve localization accuracy. However, some studies show that inter-class method call type faults account for more than 20%, which means such methods still have certain limitations. To solve the above problems, this paper proposes a two-phase software fault localization based on relational graph convolutional neural networks (Two-RGCNFL). Firstly, in Phase 1, the method call dependence graph (MCDG) of the program is constructed, the intra-class and inter-class dependencies in MCDG are extracted by using the relational graph convolutional neural network, and the classifier is used to identify the faulty methods. Then, the GraphSMOTE algorithm is improved to alleviate the impact of class imbalance on classification accuracy. Aiming at the problem of parallel ranking of element suspicious values in traditional SBFL technology, in Phase 2, Doc2Vec is used to learn static features, while spectrum information serves as dynamic features. A RankNet model based on siamese multi-layer perceptron is constructed to score and rank statements in the faulty method. This work conducts experiments on 5 real projects of Defects4J benchmark. Experimental results show that, compared with the traditional SBFL technique and two baseline methods, our approach improves the Top-1 accuracy by 262.86%, 29.59% and 53.01%, respectively, which verifies the effectiveness of Two-RGCNFL. Furthermore, this work verifies the importance of inter-class dependencies through ablation experiments.展开更多
文摘Steel surface defect detection is a key part of the production process in the steel industry.The traditional manual inspection methods are inefficient and costly.With the rapid development of deep learning technology,automatic detection and segmentation of steel surface defects based on deep neural networks has received widespread attention and demonstrated good performance in several real-world scenarios.However,challenges remain due to the obvious inter-class similarity and intra-class variation problems in steel defect images,impacting accuracy and robustness.This paper systematically summarizes the representative researches on solving the inter-class similarity and intra-class variation problems in recent years,and focuses on analyzing the innovations of different methods in network structure design.In addition,this paper also discusses the shortcomings of the current research,and proposes a new idea of fusion of mainstream modeling method.By employing a subcomparison module to enhance feature similarity within classes and differentiate across classes,followed by pyramid feature fusion to optimize computational efficiency,this study aims to advance high-precision intelligent recognition of steel surface defects.This study reveals that the approach not only addresses existing challenges but also provides a foundation for future advancements in steel defect detection technologies.
基金funded by the Youth Fund of the National Natural Science Foundation of China(Grant No.42261070).
文摘Spectrum-based fault localization (SBFL) generates a ranked list of suspicious elements by using the program execution spectrum, but the excessive number of elements ranked in parallel results in low localization accuracy. Most researchers consider intra-class dependencies to improve localization accuracy. However, some studies show that inter-class method call type faults account for more than 20%, which means such methods still have certain limitations. To solve the above problems, this paper proposes a two-phase software fault localization based on relational graph convolutional neural networks (Two-RGCNFL). Firstly, in Phase 1, the method call dependence graph (MCDG) of the program is constructed, the intra-class and inter-class dependencies in MCDG are extracted by using the relational graph convolutional neural network, and the classifier is used to identify the faulty methods. Then, the GraphSMOTE algorithm is improved to alleviate the impact of class imbalance on classification accuracy. Aiming at the problem of parallel ranking of element suspicious values in traditional SBFL technology, in Phase 2, Doc2Vec is used to learn static features, while spectrum information serves as dynamic features. A RankNet model based on siamese multi-layer perceptron is constructed to score and rank statements in the faulty method. This work conducts experiments on 5 real projects of Defects4J benchmark. Experimental results show that, compared with the traditional SBFL technique and two baseline methods, our approach improves the Top-1 accuracy by 262.86%, 29.59% and 53.01%, respectively, which verifies the effectiveness of Two-RGCNFL. Furthermore, this work verifies the importance of inter-class dependencies through ablation experiments.
