17 April 2024 in IEEE Transactions on Instrumentation and Measurement
Tiny defect detection is a knotty task in industrial electronics production. Existing traditional and deep learning methods have achieved satisfactory performance, however, they still face challenges in accuracy, generalization ability, and computational complexity. Therefore, this study designs a tiny defect detection-based you only look once (TDD-YOLO) model and proposes an innovative compression training strategy to train on low-resolution images and test on original images. First, a four-ME layers structure is adopted to the backbone network, to integrate more underlying information and extract effective features. In addition, a miniature detection head is incorporated into the head network to improve the accuracy and generalization performance of you only look once (YOLO). Meanwhile, TDD-YOLO introduces wise intersection over union (W-IoU) to reevaluate the loss of bounding box regression (BBR) and reduce false negatives by fitting the model well to regular quality anchor boxes. Finally, an image compression method at different ratios is applied in the proposed compression training strategy, to reduce computational complexity and surprisingly further improve accuracy. Comprehensive experiments on several variable compressed datasets which are based on a public printed circuit board (PCB) defect dataset validate the effectiveness of our theoretical approach and illustrate that our proposed method outperforms state-of-the-art methods.
25 August 2025 in IEEE Transactions on Instrumentation and Measurement
Steel plate fault detection is a knotty task in the manufacturing industry. Existing methods have achieved satisfactory performance, however, they still face challenges in accuracy, generalization ability, and capturing important features. Therefore, this study designs a steel plate fault detection method based on an Improved Set transformer and Kolmogorov-Arnold Network (ISet transformer-KAN). Firstly, a feature importance evaluation mechanism is incorporated into the Set transformer, dynamically optimizing seed vectors according to input data to better capture fault-related features. This modification improves information aggregation while leveraging the Set transformer’s inherent permutation invariance, ensuring stability regardless of feature order. Furthermore, KAN employs learnable activation functions with grid expansion mechanisms to progressively capture subtle feature details in steel plate faults, improving generalization capability and detection accuracy. Finally, to further validate the effectiveness of the proposed method, comprehensive experiments were conducted using the Steel Plate Faults dataset from the UCI Machine Learning Repository. The experimental results demonstrate that the proposed method outperforms traditional machine learning methods by an average of 10%-15% in accuracy and modern deep learning approaches by an average of 2%-3% in accuracy. The implementation code supporting the findings of this study is available at https://github.com/123fggv/maincode.