HL-YOLO: Insulator Defect Detection Based on Bidirectional Feature Fusion

Fengyang Han; Hongbo Bi; Wendi Yan; Rui Dai; Jianjun Xu

doi:10.54097/79qxwq70

Authors

Fengyang Han
Hongbo Bi
Wendi Yan
Rui Dai
Jianjun Xu

DOI:

https://doi.org/10.54097/79qxwq70

Keywords:

YOLOv8, Object Detective, Power Transmission Lines.

Abstract

This paper proposes a Self-attention Feature Fusion (SFF) module to enhance multi-scale defect detection in power transmission lines. The SFF module integrates the Convolutional Block Attention Module (CBAM) into the Dual Attention Transformer (DuAT) framework, enabling adaptive fusion of high-resolution and low-resolution features. By recalibrating feature weights across hierarchical levels, SFF significantly improves feature extraction for small-scale objects and fine-grained anomalies while simplifying the model architecture.

References

[1]Zhang S, Zhang X, Wan S, Ren W, Zhao L, and Shen L, "Generative adversarial and self-supervised dehazing network," in IEEE Trans. Ind. Informat., vol. 20, no. 3, pp. 4187–4197, Mar. 2024.

[2]LeCun Y, Bengio Y, and Hinton G, "Deep learning," in Nature, vol. 521, no. 7553, pp. 436–444, May 2015.

[3]Woo S, Park J, Lee J Y, "Cbam: Convolutional block attention module"[C]//Proceedings of the European conference on computer vision (ECCV). 2018: 3-19.

[4]Tang F, Xu Z, Huang Q, Wang J, Hou X, Su J, Liu J, "DuAT: Dual-Aggregation Transformer Network for Medical Image Segmentation, " in Chinese Conference on Pattern Recognition and Computer Vision (PRCV). Singapore: Springer Nature Singapore, 2023: 343-356.

[5]Chen J, Fu Z, Cheng X, and Wang F, "An method for power lines insulator defect detection with attention feedback and double spatial pyramid," in Electr. Power Syst. Res., vol. 218, May 2023, Art. no. 109175.

[6]Wu M, Guo L, Chen R, Du W, Wang J, Liu M, Kong X, and Tang J, "Improved YOLOX foreign object detection algorithm for transmission lines," in Wireless Commun. Mobile Comput., vol. 2022, pp. 1–10, Oct. 2022.

[7]Liu W, Anguelov D, Erhan D, Szegedy C, Reed S, Fu C, "Ssd: Single shot multi-box detector" in Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. Springer International Publishing, 2016: 21-37.

[8]Ren S, He K, Girshick R, and Sun J, “Faster R-CNN: Towards real-time object detection with region proposal networks,” in Proc. Adv. Neural Inf.Process. Syst., vol. 39, no. 6, pp. 1137–1149, 2016.

[9]Ge Z, Liu S, Wang F, Li Z, Sun S,"Yolox: Exceeding yolo series in 2021," in arXiv preprint arXiv:2107.08430.

[10]Wang A, Chen H, Liu L, Chen K," Yolov10: Real-time end-to-end object detection," in Advances in Neural Information Processing Systems, 2024, 37: 107984-108011.