EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

LFODet: Lightweight Few-Shot Object Detection with Meta-Learning in Remote Sensing Images

Haoran Wu iD, Xuan Fang, Haonan Xiong, Xiaomei Yang iD

Sensors · 09.07.2026 · 26 (14) · 4371

DOI: 10.3390/s26144371 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 09.07.2026. Verfasst von Haoran Wu und 3 weiteren Autor:innen. Beteiligte Einrichtung: College of Electrical Engineering, Sichuan University, Chengdu 610065, China. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Balancing detection accuracy with model lightweightness remains a key challenge in remote sensing object detection. Although convolutional neural networks have improved performance, they typically require large-scale datasets, making few-shot detection of novel classes difficult. To tackle this, we propose LFODet, a lightweight few-shot object detection network based on meta-learning. It uses two parallel branches to rapidly adapt to novel classes with limited samples while maintaining performance on base classes. For efficient feature representation, we integrate Semantic Ghost Channel Attention (GCA) and Fine-Grained Ghost Spatial Attention (GSA) to enhance semantic discriminability and spatial detail preservation. Moreover, we leverage Ghost convolutions to reduce computational complexity. The model is trained in three stages: base-class pre-training, meta-learner optimization, and few-shot fine-tuning. Experiments on DIOR and NWPU VHR-10 demonstrate that LFODet achieves stable and balanced performance across various few-shot learning scenarios. As validated on these benchmark datasets, this work provides a practical solution for resource-constrained remote sensing applications requiring rapid adaptation to new targets.

Schlagwörter

Stichwörter der Autor:innen

few-shot detectionlightweight convolutional neural network (CNN)meta-learningobject detectionremote sensing images (RSIs)

Autor:innen und Institutionen

  • College of Electrical Engineering, Sichuan University, Chengdu 610065, China.
  • Key Laboratory of Information and Automation Technology in Sichuan Province, Sichuan University, Chengdu 610065, China.

ORCID-Profile: Haoran Wu, Xiaomei Yang

Zitieren

APA

Wu, H., Fang, X., Xiong, H., & Yang, X. (2026). LFODet: Lightweight Few-Shot Object Detection with Meta-Learning in Remote Sensing Images. Sensors, 26(14), 4371. https://doi.org/10.3390/s26144371

Vancouver

Wu H, Fang X, Xiong H, Yang X. LFODet: Lightweight Few-Shot Object Detection with Meta-Learning in Remote Sensing Images. Sensors. 2026;26(14):4371. doi:10.3390/s26144371

BibTeX
@article{Wu2026,
  author = {Wu, Haoran and Fang, Xuan and Xiong, Haonan and Yang, Xiaomei},
  title = {LFODet: Lightweight Few-Shot Object Detection with Meta-Learning in Remote Sensing Images},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {14},
  pages = {4371},
  doi = {10.3390/s26144371}
}
Kostenlos starten →