EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

HL-YOLO: Improving Vehicle Damage Detection with Heterogeneous Convolutions and Large-Kernel Attention

Weijun Li, Huawei Xie, Peiteng Lin, Liyan Huang

World Electric Vehicle Journal · 21.11.2025 · 16 (12) · 640

DOI: 10.3390/wevj16120640 ↗

Kurz gefasst

Fachartikel in der Zeitschrift World Electric Vehicle Journal, online erschienen am 21.11.2025. Verfasst von Weijun Li und 3 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Innovation of Policing Science and Technology, Fujian province, Science and Technology Special Project for Provincial Units in 2023 from the Department of Finance of Fujian Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate vehicle damage detection is essential in intelligent transportation systems, insurance claim assessment, and automotive maintenance. Although conventional detection models demonstrate strong performance, they still struggle to capture fine-grained details and long-range dependencies, which can constrain their effectiveness in real-world applications. To address these limitations, we propose HL-YOLO, an enhanced YOLO11-based architecture that integrates Heterogeneous Convolutions (HetConv) to improve feature extraction diversity and Large-Kernel Attention (LSKA) to strengthen contextual representation. Model evaluation results on a vehicle damage dataset demonstrate that HL-YOLO consistently outperforms the YOLO11 baseline, achieving relative improvements of 2.5% in precision, 5.8% in recall, 3.9% in mAP50, and 3.1% in mAP50–95. These results underscore the model’s robustness in identifying complex damage types, ranging from scratches and dents to accident-induced damage. Although inference latency increased moderately due to the added architectural complexity, the overall accuracy gains confirm the effectiveness of HL-YOLO in scenarios where detection reliability is prioritized over real-time speed. The proposed model shows strong potential for deployment in insurance automation, intelligent traffic monitoring, and vehicle after-service systems, providing a reliable framework for accurate vehicle damage assessment.

Förderung

  • Innovation of Policing Science and Technology, Fujian province
  • Science and Technology Special Project for Provincial Units in 2023 from the Department of Finance of Fujian Province

Zitieren

APA

Li, W., Xie, H., Lin, P., & Huang, L. (2025). HL-YOLO: Improving Vehicle Damage Detection with Heterogeneous Convolutions and Large-Kernel Attention. World Electric Vehicle Journal, 16(12), 640. https://doi.org/10.3390/wevj16120640

Vancouver

Li W, Xie H, Lin P, Huang L. HL-YOLO: Improving Vehicle Damage Detection with Heterogeneous Convolutions and Large-Kernel Attention. World Electric Vehicle Journal. 2025;16(12):640. doi:10.3390/wevj16120640

BibTeX
@article{Li2025,
  author = {Li, Weijun and Xie, Huawei and Lin, Peiteng and Huang, Liyan},
  title = {HL-YOLO: Improving Vehicle Damage Detection with Heterogeneous Convolutions and Large-Kernel Attention},
  journal = {World Electric Vehicle Journal},
  year = {2025},
  volume = {16},
  number = {12},
  pages = {640},
  doi = {10.3390/wevj16120640}
}
Kostenlos starten →