Wheel-tread defect recognition is a crucial step in ensuring the safety of train wheel-rail system services. However, the diverse and complex nature of wheel-tread defects, coupled with the presence of minor defect features, poses significant challenges in accurately identifying defects using existing deep convolutional neural networks. To address this problem, we developed a small target defect-detection module and proposed a railway wheel-tread defect-recognition method based on an improved convolutional neural network. First, a deformable convolutional attention-enhanced bottleneck module was designed to achieve an adaptive adjustment of the network receptive field in the backbone network. Second, an adaptive spatial and channel enhancement module was constructed to further improve the network’s sensitivity and processing capabilities for different features. Third, we designed a new module called the spatial grouped attention fusion pyramid module to enhance the extraction and fusion capabilities of multiscale features through grouping and fusion of spatial attention mechanisms, enabling the effective extraction and discrimination of defect multi-layer semantic features. Finally, experiments were conducted using a tread defect dataset with an imbalance ratio of 10:1. The experimental results demonstrated the excellent performance of the proposed model on public datasets. The average mAP@0.5 value increased from 90.6% to 91.9%. Similarly, the observed average mAP@0.5:0.95 value increased from 53.9% to 54.9%.

1. [1] H. Yang, J. He, Z. Liu, and C. Zhang, “LLD-MFCOS: A multiscale anchor-free detector based on label localization distillation for wheelset tread defect detection,” IEEE Trans. Instrum. Meas., Vol.73, Article No.5003815, 2024. https://doi.org/10.1109/TIM.2023.3316214
2. [2] Z. Huang et al., “Joining spatial deformable convolution and a dense feature pyramid for surface defect detection,” IEEE Trans. Instrum. Meas., Vol.73, Article No.5012614, 2024. https://doi.org/10.1109/TIM.2024.3370962
3. [3] Z. Li, B. Li, S. G. Jahng, and C. Jung, “Improved VGG algorithm for visual prosthesis image recognition,” IEEE Access, Vol.12, pp. 45727-45739, 2024. https://doi.org/10.1109/ACCESS.2024.3380839
4. [4] C. Zhang, Y. Xu, Z. Sheng, J. He, and L. Yin, “Deformable residual attention network for defect detection of train wheelset tread,” Vis. Comput., Vol.40, No.3, pp. 1775-1785, 2024. https://doi.org/10.1007/s00371-023-02885-z
5. [5] X. Qian, X. Wang, S. Yang, and J. Lei, “LFF-YOLO: A YOLO algorithm with lightweight feature fusion network for multi-scale defect detection,” IEEE Access, Vol.10, pp. 130339-130349, 2022. https://doi.org/10.1109/ACCESS.2022.3227205
6. [6] H. Dong, Y. Li, and R. Liu, “A detection algorithm based on improved cascade R-CNN for UAV aerial images,” IEEE 3rd Int. Conf. Electron. Technol. Commun. Inf., pp. 700-704, 2023. https://doi.org/10.1109/ICETCI57876.2023.10176806
7. [7] L. Yang et al., “An improving Faster-RCNN with multi-attention ResNet for small target detection in intelligent autonomous transport with 6G,” IEEE Trans. Intell. Transp. Syst., Vol.24, No.7, pp. 7717-7725, 2023. https://doi.org/10.1109/TITS.2022.3193909
8. [8] C. Peng, X. Li, and Y. Wang, “TD-YOLOA: An efficient YOLO network with attention mechanism for tire defect detection,” IEEE Trans. Instrum. Meas., Vol.72, Article No.3529111, 2023. https://doi.org/10.1109/TIM.2023.3312753
9. [9] S. J. Li, Y. X. Liu, M. Li, and L. Ding, “DF-YOLO: Highly accurate transmission line foreign object detection algorithm,” IEEE Access, Vol.11, pp. 108398-108406, 2023. https://doi.org/10.1109/ACCESS.2023.3321385
10. [10] J. Liu, D. Du, J. He, and C. Zhang, “Prediction of remaining useful life of railway tracks based on DMGDCC-GRU hybrid model and transfer learning,” IEEE Trans. Veh. Technol., Vol.73, No.6, pp. 7561-7575, 2024. https://doi.org/10.1109/TVT.2024.3351763
11. [11] T. Ye et al., “Real-time object detection network in UAV-vision based on CNN and transformer,” IEEE Trans. Instrum. Meas., Vol.72, Article No.2505713, 2023. https://doi.org/10.1109/TIM.2023.3241825
12. [12] Q. Ding et al., “CF-YOLO: Cross fusion YOLO for object detection in adverse weather with a high-quality real snow dataset,” IEEE Trans. Intell. Transp. Syst., Vol.24, No.10, pp. 10749-10759, 2023. https://doi.org/10.1109/TITS.2023.3285035
13. [13] H. Tan, Z. Zhang, Y. Li, X. Shi, and F. Zhou, “Multi-scale feature fusion and distribution similarity network for few-shot automatic modulation classification,” IEEE Signal Process. Lett., Vol.31, pp. 2890-2894, 2024. https://doi.org/10.1109/LSP.2024.3470762
14. [14] Q. Zhang, J. Zhang, Y. Li, C. Zhu, and G. Wang, “IL-YOLO: An efficient detection algorithm for insulator defects in complex backgrounds of transmission lines,” IEEE Access, Vol.12, pp. 14532-14546, 2024. https://doi.org/10.1109/ACCESS.2024.3358205
15. [15] M. Qian, Y.-F. Li, and H. Wu, “Deep imbalanced separation network: A holistic fault detection framework considering class-imbalance and partial label-unknown,” IEEE Trans. Ind. Inform., Vol.20, No.11, pp. 13026-13035, 2024. https://doi.org/10.1109/TII.2024.3431048
16. [16] L. Zhao et al., “MSFA-YOLO: A multi-scale SAR ship detection algorithm based on fused attention,” IEEE Access, Vol.12, pp. 24554-24568, 2024. https://doi.org/10.1109/ACCESS.2024.3365777
17. [17] X. Li, W. Diao, Y. Mao, X. Li, and X. Sun, “SCLNet: A scale-robust complementary learning network for object detection in UAV images,” IEEE Trans. Geosci. Remote Sens., Vol.62, Article No.5651119, 2024. https://doi.org/10.1109/TGRS.2024.3505425
18. [18] Y. Zhang et al., “Improved YOLOv8 insulator fault detection algorithm based on BiFormer,” IEEE 5th Int. Conf. Power Intell. Comput. Syst., pp. 962-965, 2023. https://doi.org/10.1109/ICPICS58376.2023.10235397
19. [19] L.-Q. Zhang, Z.-T. Liu, and C.-S. Jiang, “An improved SimAM based CNN for facial expression recognition,” 41st Chin. Control Conf., pp. 6510-6514, 2022. https://doi.org/10.23919/CCC55666.2022.9902045
20. [20] D. Zhan et al., “Interactive fusion distillation network embedded with Poisson images for wheel tread defect detection,” IEEE Trans. Instrum. Meas., Vol.74, Article No.2522413, 2025. https://doi.org/10.1109/TIM.2025.3551489
21. [21] Y. Liu, H. Yang, and C. Wu, “Unveiling patterns: A study on semi-supervised classification of strip surface defects,” IEEE Access, Vol.11, pp. 119933-119946, 2023. https://doi.org/10.1109/ACCESS.2023.3326843
22. [22] K. Kim et al., “UAV chasing based on YOLOv3 and object tracker for counter UAV systems,” IEEE Access, Vol.11, pp. 34659-34673, 2023. https://doi.org/10.1109/ACCESS.2023.3264603
23. [23] Z. Lu, L. Ding, Z. Wang, L. Dong, and Z. Guo, “Road condition detection based on deep learning YOLOv5 network,” IEEE 3rd Int. Conf. Electron. Technol. Commun. Inf., pp. 497-501, 2023. https://doi.org/10.1109/ICETCI57876.2023.10176545