Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
166 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Efficient Visual Fault Detection for Freight Train via Neural Architecture Search with Data Volume Robustness (2405.17004v1)

Published 27 May 2024 in cs.CV and eess.IV

Abstract: Deep learning-based fault detection methods have achieved significant success. In visual fault detection of freight trains, there exists a large characteristic difference between inter-class components (scale variance) but intra-class on the contrary, which entails scale-awareness for detectors. Moreover, the design of task-specific networks heavily relies on human expertise. As a consequence, neural architecture search (NAS) that automates the model design process gains considerable attention because of its promising performance. However, NAS is computationally intensive due to the large search space and huge data volume. In this work, we propose an efficient NAS-based framework for visual fault detection of freight trains to search for the task-specific detection head with capacities of multi-scale representation. First, we design a scale-aware search space for discovering an effective receptive field in the head. Second, we explore the robustness of data volume to reduce search costs based on the specifically designed search space, and a novel sharing strategy is proposed to reduce memory and further improve search efficiency. Extensive experimental results demonstrate the effectiveness of our method with data volume robustness, which achieves 46.8 and 47.9 mAP on the Bottom View and Side View datasets, respectively. Our framework outperforms the state-of-the-art approaches and linearly decreases the search costs with reduced data volumes.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (38)
  1. L. Butera, A. Ferrante, M. Jermini, M. Prevostini, and C. Alippi, “Precise agriculture: Effective deep learning strategies to detect pest insects,” IEEE-CAA J. Automatica Sin., vol. 9, no. 2, pp. 246–258, 2021.
  2. Y. Zhang, Y. Zhou, H. Pan, B. Wu, and G. Sun, “Visual fault detection of multi-scale key components in freight trains,” IEEE Trans. Ind. Informat., vol. 19, pp. 9082–9090, 2022.
  3. S. R. Saufi, Z. A. B. Ahmad, M. S. Leong, and M. H. Lim, “Gearbox fault diagnosis using a deep learning model with limited data sample,” IEEE Trans. Ind. Informat., vol. 16, no. 10, pp. 6263–6271, 2020.
  4. J. Sun, Y. Xie, and X. Cheng, “A fast bolt-loosening detection method of running train’s key components based on binocular vision,” IEEE Access, vol. 7, pp. 32227–32239, 2019.
  5. C. Chen, X. Zou, Z. Zeng, Z. Cheng, L. Zhang, and S. C. H. Hoi, “Exploring structural knowledge for automated visual inspection of moving trains,” IEEE Trans. Cybern., vol. 52, no. 2, pp. 1233–1246, 2022.
  6. Z. Zhou, Y. Hu, X. Deng, D. Huang, and Y. Lin, “Fault detection of train height valve based on nanodet-resnet101,” in YAC, pp. 709–714, 2021.
  7. G. Ghiasi, T. Lin, and Q. V. Le, “Nas-fpn: Learning scalable feature pyramid architecture for object detection,” in CVPR, pp. 7036–7045, 2019.
  8. N. Wang, Y. Gao, H. Chen, P. Wang, Z. Tian, C. Shen, and Y. Zhang, “Nas-fcos: Fast neural architecture search for object detection,” in CVPR, 2020.
  9. Y. Chen, T. Yang, X. Zhang, G. Meng, X. Xiao, and J. Sun, “Detnas: Backbone search for object detection,” in NeurIPS, pp. 6638–6648, 2019.
  10. T. Liang, Y. Wang, Z. Tang, G. Hu, and H. Ling, “OPANAS: one-shot path aggregation network architecture search for object detection,” in CVPR, pp. 10195–10203, 2021.
  11. Y. Sun, N. Qin, and L. Ma, “High-speed train bogie faults diagnosis using singular spectrum analysis,” in DDCLS, pp. 56–59, 2018.
  12. X. Zhou, D. Wang, and P. Krähenbühl, “Objects as Points,” arXiv preprint arXiv:1904.07850, 2019.
  13. Z. Dong, G. Li, Y. Liao, F. Wang, P. Ren, and C. Qian, “Centripetalnet: Pursuing high-quality keypoint pairs for object detection,” in CVPR, pp. 10519–10528, 2020.
  14. Z. Tian, C. Shen, H. Chen, and T. He, “FCOS: Fully Convolutional One-Stage Object Detection,” in ICCV, pp. 9626–9635, 2019.
  15. X. Li, W. Wang, L. Wu, S. Chen, X. Hu, J. Li, J. Tang, and J. Yang, “Generalized focal loss: Learning qualified and distributed bounding boxes for dense object detection,” NeurIPS, vol. 33, pp. 21002–21012, 2020.
  16. T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, “Focal loss for dense object detection,” in ICCV, pp. 2980–2988, 2017.
  17. Q. Chen, Y. Wang, T. Yang, X. Zhang, J. Cheng, and J. Sun, “You only look one-level feature,” in CVPR, pp. 13039–13048, 2021.
  18. Z. Ge, S. Liu, F. Wang, Z. Li, and J. Sun, “Yolox: Exceeding yolo series in 2021,” arXiv preprint arXiv:2107.08430, 2021.
  19. C.-Y. Wang, A. Bochkovskiy, and H.-Y. M. Liao, “Yolov7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors,” CVPR, pp. 7464–7475, 2022.
  20. G. Jocher, A. Chaurasia, and J. Qiu, “Ultralytics yolov8,” 2023.
  21. S. Ren, K. He, R. Girshick, and J. Sun, “Faster r-cnn: Towards real-time object detection with region proposal networks,” TPAMI, vol. 39, no. 06, pp. 1137–1149, 2017.
  22. J. Pang, K. Chen, J. Shi, H. Feng, W. Ouyang, and D. Lin, “Libra R-CNN: towards balanced learning for object detection,” in CVPR, pp. 821–830, 2019.
  23. P. Sun, R. Zhang, Y. Jiang, T. Kong, C. Xu, W. Zhan, M. Tomizuka, L. Li, Z. Yuan, C. Wang, and P. Luo, “Sparse R-CNN: End-to-End Object Detection with Learnable Proposals,” in CVPR, pp. 14449–14458, 2021.
  24. S. Yang, L. Xu, M. Zhou, X. Yang, J. Yang, and Z. Huang, “Skill-transferring knowledge distillation method,” IEEE Trans. Circuits Syst., vol. 33, pp. 6487–6502, 2023.
  25. Z. Huang, S. Yang, M. Zhou, Z. Li, Z. Gong, and Y. Chen, “Feature map distillation of thin nets for low-resolution object recognition,” IEEE Transactions on Image Process., vol. 31, pp. 1364–1379, 2022.
  26. X. Zhu, W. Su, L. Lu, B. Li, X. Wang, and J. Dai, “Deformable detr: Deformable transformers for end-to-end object detection,” in ICLR, 2021.
  27. A. Xu, A. Yao, A. Li, A. Liang, and A. Zhang, “Auto-fpn: Automatic network architecture adaptation for object detection beyond classification,” in ICCV, pp. 6649–6658, 2019.
  28. Y. Zhong, Z. Deng, S. Guo, M. R. Scott, and W. Huang, “Representation sharing for fast object detector search and beyond,” in ECCV, pp. 471–487, 2020.
  29. K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014.
  30. H. Liu, K. Simonyan, and Y. Yang, “Darts: Differentiable architecture search,” arXiv preprint arXiv:1806.09055, 2018.
  31. T.-Y. Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Dollár, and C. L. Zitnick, “Microsoft coco: Common objects in context,” in ECCV, pp. 740–755, 2014.
  32. M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, “Mobilenetv2: Inverted residuals and linear bottlenecks,” in CVPR, pp. 4510–4520, 2018.
  33. Y. Zhang, M. Liu, Y. Yang, Y. Guo, and H. Zhang, “A unified light framework for real-time fault detection of freight train images,” IEEE Trans. Ind. Informat., vol. 17, no. 11, pp. 7423–7432, 2021.
  34. K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in CVPR, pp. 770–778, 2016.
  35. K. He, X. Zhang, S. Ren, and J. Sun, “Identity mappings in deep residual networks,” in ECCV, pp. 630–645, 2016.
  36. Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” in ICCV, pp. 10012–10022, 2021.
  37. C.-Y. Wang, H.-Y. M. Liao, Y.-H. Wu, P.-Y. Chen, J.-W. Hsieh, and I.-H. Yeh, “Cspnet: A new backbone that can enhance learning capability of cnn,” in CVPR, pp. 390–391, 2020.
  38. H. Law, Y. Teng, O. Russakovsky, and J. Deng, “Cornernet: Detecting objects as paired keypoints,” arXiv preprint arXiv:1808.01244, 2018.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets