A Third-Party Construction Detection Model Incorporating Residual Scaling and Channel-Adaptive Gating

Jing Li; Weixiang Tao; Xuanpeng Liu; Yanlei Ye; Ran Wang; Haiquan Jin

doi:10.54097/ctrdag18

Authors

Jing Li Zhengyuan Geographic Information Group Co., LTD, Beijing 101300, China
Weixiang Tao Zhengyuan Geographic Information Group Co., LTD, Beijing 101300, China
Xuanpeng Liu Zhengyuan Geographic Information Group Co., LTD, Beijing 101300, China
Yanlei Ye Zhengyuan Geographic Information Group Co., LTD, Beijing 101300, China
Ran Wang Beijing Zhengyuan Digital Technology Information Co., LTD, Beijing 101300, China
Haiquan Jin Beijing Zhengyuan Digital Technology Information Co., LTD, Beijing 101300, China

DOI:

https://doi.org/10.54097/ctrdag18

Keywords:

Third-party construction detection; End-to-end object detection; Detection Transformer (DETR); Real-time monitoring.

Abstract

Third-party construction is a leading cause of urban gas pipeline damage. Unregulated digging and drilling frequently rupture pipes, disrupt service, and create safety hazards. To address this, we propose TPC-DETR—an end-to-end detector built on the DETR framework—that identifies risky construction activities near pipelines and helps prevent human-caused damage. Three enhancements improve the baseline. In the backbone, we add a Residual Scaled Block (RSB) that stabilizes deep-network training and strengthens feature representation, with zero extra parameters or FLOPs. Between the backbone and neck, we insert a Channel-Adaptive Cascaded Gating Layer (CACG-Layer) that lowers computational cost while boosting speed and accuracy through fine-grained multi-branch feature modulation. In the feature extraction path, we adopt re-parameterized RefConv, whose kernels adjust dynamically at inference based on learned weights—improving adaptability and generalization at no additional cost. On our self-built dataset, TPC-DETR achieves 97% mAP@0.5 and 81.6% mAP@0.5:0.95. These results show that the model reliably detects a wide range of third-party construction activities in complex pipeline environments, and is practical for real-world deployment.

Downloads

Download data is not yet available.

References

[1] Hong, B., Li, X., Song, S., & Liu, Y. (2020). Optimal planning and modular infrastructure dynamic allocation for shale gas production. Applied Energy, 261, 114439. https://doi.org/10.1016/j.apenergy.2019.114439 DOI: https://doi.org/10.1016/j.apenergy.2019.114439

[2] Koroteev, D., & Tekic, Z. (2021). Artificial intelligence in oil and gas upstream: Trends, challenges, and scenarios for the future. Energy and AI, 3, 100041. https://doi.org/10.1016/j.egyai.2020.100041 DOI: https://doi.org/10.1016/j.egyai.2020.100041

[3] Witek, M., & Uilhoorn, F. E. (2021). Influence of gas transmission network failure on security of supply. Journal of Natural Gas Science and Engineering, 90, 103877. https://doi.org/10.1016/j.jngse.2021.103877 DOI: https://doi.org/10.1016/j.jngse.2021.103877

[4] Rusin, A., Stolecka Antczak, K., Kapusta, K., & Grzelak, M. (2021). Analysis of the effects of failure of a gas pipeline caused by a mechanical damage. Energies, 14(22), 7686. https://doi.org/10.3390/en14227686 DOI: https://doi.org/10.3390/en14227686

[5] Hussain, R., Lee, D., Abbas, M. S., & Ali, S. (2026). Vision language model based intelligent assistant for onsite construction safety inspection. Automation in Construction, 182, 106728. https://doi.org/10.1016/j.autcon.2025.106728 DOI: https://doi.org/10.1016/j.autcon.2025.106728

[6] Kim, J., Gong, Y., Chi, S., & Park, J. (2026). Towards transparent object detection models for construction sites: explainable AI and error classification. Advanced Engineering Informatics, 71, 104245. https://doi.org/10.1016/j.aei.2025.104245 DOI: https://doi.org/10.1016/j.aei.2025.104245

[7] Awuku, B., Huang, Y., & Yodo, N. (2023). Predicting natural gas pipeline failures caused by natural forces: an artificial intelligence classification approach. Applied Sciences, 13(7), 4322. https://doi.org/10.3390/app13074322 DOI: https://doi.org/10.3390/app13074322

[8] Li, L., Cui, X., Wang, J., & Zhang, H. (2026). Automated Tracking of Worker and Heavy Equipment on Tunnel Construction Sites: Deep Learning Framework. Journal of Construction Engineering and Management, 152(2), 04025237. https://doi.org/10.1061/JCEMD4.COENG 16807 DOI: https://doi.org/10.1061/JCEMD4.COENG-16807

[9] Vedaldi, A., & Zisserman, A. (2016). Vgg convolutional neural networks practical. University of Oxford.

[10] He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770 778). https://doi.org/10.1109/CVPR.2016.90 DOI: https://doi.org/10.1109/CVPR.2016.90

[11] Ouyang, S., Wang, H., Niu, Z., & Li, Y. (2023). HSVLT: Hierarchical scale aware vision language transformer for multi label image classification. In Proceedings of the 31st ACM International Conference on Multimedia (pp. 4768 4777). https://doi.org/10.1145/3581783.3612159 DOI: https://doi.org/10.1145/3581783.3612159

[12] Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 779 788). https://doi.org/10.48550/arXiv.1506.02640 DOI: https://doi.org/10.1109/CVPR.2016.91

[13] Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C. Y., & Berg, A. C. (2016). Ssd: Single shot multibox detector. In B. Leibe, J. Matas, N. Sebe, & M. Welling (Eds.), Computer Vision–ECCV 2016 (pp. 21 37). Springer International Publishing. https://doi.org/10.1007/978 3 319 46448 0_2 DOI: https://doi.org/10.1007/978-3-319-46448-0_2

[14] Carion, N., Massa, F., Synnaeve, G., Usunier, N., Kirillov, A., & Zagoruyko, S. (2020). End to end object detection with transformers. In A. Vedaldi, H. Bischof, T. Brox, & J. M. Frahm (Eds.), Computer Vision – ECCV 2020 (pp. 213 229). Springer International Publishing. https://doi.org/10.1007/978 3 030 58452 8_13 DOI: https://doi.org/10.1007/978-3-030-58452-8_13

[15] Hu, H., Guo, F., & Liu, Z. (2023). Target Detection in Construction Scenes Using an Improved YOLOX S Model. Journal of Frontiers of Computer Science & Technology, 17(5). https://doi.org/10.3778/j.issn.1673 9418.2205012

[16] Meng, L., Peng, Z., Zhou, J., & Liu, T. (2020). Real time detection of ground objects based on unmanned aerial vehicle remote sensing with deep learning: Application in excavator detection for pipeline safety. Remote Sensing, 12(1), 182. https://doi.org/10.3390/rs12010182 DOI: https://doi.org/10.3390/rs12010182

[17] Hao, F., Zhang, T., He, G., & Wang, Q. (2024). CaSnLi YOLO: construction site multi target detection method based on improved YOLOv5s. Measurement Science and Technology, 35(8), 085202. https://doi.org/10.1088/1361 6501/ad42c5 DOI: https://doi.org/10.1088/1361-6501/ad42c5

[18] Wang, W., He, W., Bai, M., & Liu, S. (2025). Improved detection and reidentification algorithm for natural gas pipeline excavator dynamic tracking based on deep learning. Measurement, 118567. https://doi.org/10.1016/j.measurement.2025.118567 DOI: https://doi.org/10.1016/j.measurement.2025.118567

[19] Xiang, Y., Zhao, J., Wu, W., & Li, B. (2022). Automatic object detection of construction workers and machinery based on improved YOLOv5. In Y. Wang (Ed.), International Conference on Green Building, Civil Engineering and Smart City (pp. 741 749). Springer Nature Singapore. https://doi.org/10.1007/978 981 19 5217 3_74 DOI: https://doi.org/10.1007/978-981-19-5217-3_74

[20] Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021). Swin transformer: Hierarchical vision transformer using shifted windows. In Proceedings of the IEEE/CVF international conference on computer vision (pp. 10012 10022). https://doi.org/10.48550/arXiv.2103.14030 DOI: https://doi.org/10.1109/ICCV48922.2021.00986

[21] Zhu, X., Su, W., Lu, L., Li, B., Wang, X., & Dai, J. (2020). Deformable detr: Deformable transformers for end to end object detection [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2010.04159

[22] Zhao, Y., Lv, W., Xu, S., & Li, J. (2024). Detrs beat yolos on real time object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 16965 16974). https://doi.org/10.48550/arXiv.2304.08069 DOI: https://doi.org/10.1109/CVPR52733.2024.01605

[23] Liu, X., Peng, H., Zheng, N., & Yang, Y. (2023). Efficientvit: Memory efficient vision transformer with cascaded group attention. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 14420 14430). https://doi.org/10.48550/arXiv.2305.07027 DOI: https://doi.org/10.1109/CVPR52729.2023.01386

[24] Cai, Z., Ding, X., Shen, Q., & Wang, Y. (2025). Refconv: Reparameterized refocusing convolution for powerful convnets. IEEE Transactions on Neural Networks and Learning Systems. https://doi.org/10.1109/tnnls.2025.3552654 DOI: https://doi.org/10.1109/TNNLS.2025.3552654

[25] Jocher, G., Chaurasia, A., Stoken, A., & Qiu, J. (2022). ultralytics/yolov5: v7.0 yolov5 sota realtime instance segmentation [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.3908559

[26] Wang, C. Y., Bochkovskiy, A., & Liao, H. Y. M. (2023). YOLOv7: Trainable bag of freebies sets new state of the art for real time object detectors. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 7464 7475). https://doi.org/10.48550/arXiv.2207.02696 DOI: https://doi.org/10.1109/CVPR52729.2023.00721

[27] Reis, D., Nguyen, H., & Choi, J. (2023). Real Time Flying Object Detection with YOLOv8 [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2305.09972

[28] Lyu, C., Zhang, W., Huang, H., Zhou, Y., Wang, Y., Liu, Y., Zhang, S., & Chen, K. (2022). Rtmdet: An empirical study of designing real time object detectors [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2212.07784

[29] Xu, X., Jiang, Y., Chen, W., Zhang, Y., & Sun, X. (2022). Damo yolo: A report on real time object detection design [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2211.15444

[30] Wang, C. Y., Yeh, I. H., & Mark Liao, H. Y. (2025). Yolov9: Learning what you want to learn using programmable gradient information. In A. Leonardis, E. Ricci, S. Roth, O. Russakovsky, T. Sattler, & G. Varol (Eds.), Computer Vision – ECCV 2024 (pp. 1 21). Springer, Cham. https://doi.org/10.1007/978 3 031 72751 1_1 DOI: https://doi.org/10.1007/978-3-031-72751-1_1

[31] Wang, A., Chen, H., Liu, L., Ding, G., Han, J., & Lin, Z. (2024). Yolov10: Real time end to end object detection. In Advances in Neural Information Processing Systems 37 (pp. 107984 108011). https://doi.org/10.52202/079017 3429 DOI: https://doi.org/10.52202/079017-3429

[32] Khanam, R., & Hussain, M. (2024). Yolov11: An overview of the key architectural enhancements [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2410.17725