EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

SFC-YOLO: An Accuracy-Enhanced and Parameter-Efficient YOLO Framework for Small Vehicle Detection in Aerial Images

Jinwei Zuo, Qiyi He, Fulin Liu, Jieying Liu iD, Minlong Hu iD, Yiguo Wang iD, Tingting Zhang iD, Yiming Yang, Qiao Lin

Sensors · 04.08.2026 · 26 (15) · 4923

DOI: 10.3390/s26154923 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 04.08.2026. Verfasst von Jinwei Zuo und 8 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science, Hubei University of Technology, Wuhan 430068, China. Die Arbeit nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hubei Provincial Natural Science Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Small vehicle detection in aerial images is important for intelligent transportation, low-altitude inspection, urban monitoring, and vision-based sensing systems. Vehicles in aerial images often occupy few pixels and are affected by complex backgrounds, shadows, viewpoint changes, weak texture, and similar class appearances, which can cause missed detections and false alarms. To address these issues, this paper proposes SFC-YOLO, an accuracy-enhanced and parameter-efficient framework based on YOLO11n. A Feature Complementary Block (FCB) is placed at the P5/32 high-level feature stage to enhance local-detail and semantic-context compensation; a Dynamic Feature Alignment Upsampling unit (DFAU) is inserted into the first P5-to-P4 upsampling path to improve content-adaptive feature alignment; and a P5-only Hidden-State Attention (HSA) module is used in the final P5 detection branch to strengthen global semantic interaction. Experiments on the VEDAI eight-class vehicle dataset show that SFC-YOLO reduces the parameter count from 2.584 M to 2.466 M while improving the five-run mean Precision from 0.602 to 0.665, Recall from 0.599 to 0.608, and mAP50 from 0.614 to 0.652. Since the GFLOPs increase from 6.3 to 8.4, SFC-YOLO should be interpreted as a parameter-reduced but not FLOP-reduced framework. The main trade-off is improved detection accuracy with fewer trainable parameters at the cost of higher theoretical computation. Five repeated experiments yield an average mAP50 of 0.6523 ± 0.0117, quantifying the run-to-run variation under the current training protocol.

Schlagwörter

Stichwörter der Autor:innen

YOLO11naerial image detectiondynamic upsamplinghidden-state attentionintelligent transportationparameter-efficient detectionsmall vehicle detection

Autor:innen und Institutionen

  • School of Computer Science, Hubei University of Technology, Wuhan 430068, China.
  • Hubei Luojia Laboratory, Wuhan 430079, China.
  • School of Computer Science, University of Nottingham Ningbo China, Ningbo 315100, China.

ORCID-Profile: Jieying Liu, Minlong Hu, Yiguo Wang, Tingting Zhang

Förderung

  • Hubei Provincial Natural Science Foundation

Zitieren

APA

Zuo, J., He, Q., Liu, F., Liu, J., Hu, M., Wang, Y., Zhang, T., Yang, Y., & Lin, Q. (2026). SFC-YOLO: An Accuracy-Enhanced and Parameter-Efficient YOLO Framework for Small Vehicle Detection in Aerial Images. Sensors, 26(15), 4923. https://doi.org/10.3390/s26154923

Vancouver

Zuo J, He Q, Liu F, Liu J, Hu M, Wang Y, et al. SFC-YOLO: An Accuracy-Enhanced and Parameter-Efficient YOLO Framework for Small Vehicle Detection in Aerial Images. Sensors. 2026;26(15):4923. doi:10.3390/s26154923

BibTeX
@article{Zuo2026,
  author = {Zuo, Jinwei and He, Qiyi and Liu, Fulin and Liu, Jieying and Hu, Minlong and Wang, Yiguo and Zhang, Tingting and Yang, Yiming and Lin, Qiao},
  title = {SFC-YOLO: An Accuracy-Enhanced and Parameter-Efficient YOLO Framework for Small Vehicle Detection in Aerial Images},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {15},
  pages = {4923},
  doi = {10.3390/s26154923}
}
Kostenlos starten →