A Two-Stage Multi-Label Sewer Defect Identification Approach Using TResNet and CSRA Attention Module
Source abstract
Abstract
To address the coexistence of multiple defect types and class imbalance in complex urban sewer scenarios, this study proposes a two-stage multi-label recognition method for sewer defect images. In the first stage, a lightweight binary convolutional neural network is used to distinguish between normal and defective images, thereby filtering out normal samples and reducing the computational burden. In the second stage, a TResNet-based multi-label classification model is developed, and a class-specific residual attention module is introduced to enhance defect-specific spatial feature representation. In addition, F1-Normal and F2-CIW are used to evaluate the recognition performance from both normal-pipeline recognition and defect-economic-impact perspectives. Experiments on the Sewer-ML dataset show that the proposed method achieves 90.88% F1-Normal and 53.93% F2-CIW, outperforming the baseline models in overall performance.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Infrastructure Maintenance and Monitoring · Urban Stormwater Management Solutions · Water Systems and Optimization
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
College of Mechanical and Electrical Engineering, Shaanxi University of Science & Technology, Xi’an 710021, China
Funding
Key Research and Development Plan of Xianyang City, Shaanxi Province
References (22)
Show source-provided references
1. Underground sewer pipe condition assessment based on convolutional neural networks10.1016/j.autcon.2019.102849 ↗
2. Imagenet classification with deep convolutional neural networks
3. Sewer damage detection from imbalanced CCTV inspection data using deep convolutional neural networks with hierarchical classification10.1016/j.autcon.2019.01.017 ↗
4. He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2016.90 ↗
5. Chen, K. (2019). Research on Detecting Method of Sewer Defects Based on Vision. [Master’s Thesis, Harbin Institute of Technology].
6. Iandola, F.N., Han, S., Moskewicz, M.W., Ashraf, K., Dally, W.J., and Keutzer, K. (2016). SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5 MB model size. arXiv.
7. Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., and Wojna, Z. (2016). Rethinking the inception architecture for computer vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2016.308 ↗
8. Automatic detection and classification of sewer defects via hierarchical deep learning10.1109/TASE.2019.2900170 ↗
9. Automated defect classification in sewer closed circuit television inspections using deep convolutional neural networks10.1016/j.autcon.2018.03.028 ↗
10. Graph Convolutional Network Detection Model for Pipeline Defects Based on Improved Label Graph
11. Haurum, J.B., and Moeslund, T.B. (2021). Sewer-ML: A multi-label sewer defect classification dataset and benchmark. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR46437.2021.01325 ↗
12. Ridnik, T., Lawen, H., Noy, A., Baruch, E.B., Sharir, G., and Friedman, I. (2021). Tresnet: High performance gpu-dedicated architecture. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, IEEE.10.1109/WACV48630.2021.00144 ↗
13. Zhu, K., and Wu, J. (2021). Residual attention: A simple but effective method for multi-label recognition. Proceedings of the IEEE/CVF International Conference on Computer Vision, IEEE.10.1109/ICCV48922.2021.00025 ↗
14. Sandler, M., Baccash, J., Zhmoginov, A., and Howard, A. (2019). Non-discriminative data or weak model? On the relative importance of data and model resolution. Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops, IEEE.10.1109/ICCVW.2019.00133 ↗
15. Zhang, R. (2019). Making Convolutional Networks Shift-Invariant Again. arXiv.
16. Bulo, S.R., Porzi, L., and Kontschieder, P. (2018). In-place activated batchnorm for memory-optimized training of DNNs. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2018.00591 ↗
17. Hu, J., Shen, L., and Sun, G. (2018). Squeeze-and-excitation networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, IEEE.10.1109/CVPR.2018.00745 ↗
18. Lin, M., Chen, Q., and Yan, S. (2013). Network in network. arXiv.
19. Brownlee, J. (2019). Probability for Machine Learning: Discover How to Harness Uncertainty with Python, Machine Learning Mastery.
20. DANSK Vand-og Spildevandsforening (2010). Fotomanualen: TV-Inspektion af Afløbsledninger, Dansk Vand og Spildevandsforening (DANVA).
21. DANSK Vand-og Spildevandsforening (2005). Fotomanualen: Beregning af Fysisk Indeks ved TV-inspektion, Dansk Vand og Spildevandsforening (DANVA).
22. Van Rijsbergen, C.J. (1979). Information Retrieval, Butterworth-Heinemann.