Detecting infrared UAVs on edge devices through lightweight instance segmentation
PLOS One · 18.08.2025 · 20 (8) · e0330074
DOI: 10.1371/journal.pone.0330074 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 18.08.2025. Verfasst von YuZhi Chen und 5 weiteren Autor:innen. Beteiligte Einrichtung: Hebei University of Architecture, Zhangjiakou, China. Hauptthemen laut MEDLINE: Infrared Rays, Unmanned Aerial Devices, Image Processing, Computer-Assisted. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Doctoral (Postdoctoral) Fund of Hebei Normal University, Hebei Provincial Department of Science and Technology.
Abstract (Originalsprache)
Motivation Infrared unmanned aerial vehicle (UAV) detection for surveillance applications faces three conflicting requirements: accurate detection of pixel-level thermal signatures, real-time processing capabilities, and deployment feasibility on resource-constrained edge devices. Current deep learning approaches typically optimize for one or two of these objectives while compromising the third.
Method
This paper presents YOLO11-AU-IR, a lightweight instance segmentation framework that addresses these challenges through three architectural innovations. First, Efficient Adaptive Downsampling (EADown) employs dual-branch processing with grouped convolutions to preserve small-target spatial features during multi-scale fusion. Second, HeteroScale Attention Network (HSAN) implements grouped multi-scale convolutions with joint channel-spatial attention mechanisms for enhanced cross-scale feature representation. These architectural optimizations collectively reduce computational requirements while maintaining detection accuracy. Third, Adaptive Threshold Focal Loss (ATFL) introduces epoch-adaptive parameter tuning to address the extreme foreground-background imbalance inherent in infrared UAV imagery.
Results
YOLO11-AU-IR is evaluated on the AUVD-Seg300 dataset, achieving 97.7% mAP@0.50 and 75.2% mAP@0.50:0.95, surpassing the YOLO11n-seg baseline by 1.7% and 4.4%, respectively. The model reduces parameters by 24.5% and GFLOPs by 11.8% compared to YOLO11n-seg, while maintaining real-time inference at 59.8 FPS on an NVIDIA RTX 3090 with low variance. On the NVIDIA Jetson TX2, under INT8 CPU-only deployment, YOLO11-AU-IR retains 95% mAP@0.50 with minimal memory footprint and stable performance, demonstrating its practical edge compatibility. Ablation studies further confirm the complementary contributions of EADown, HSAN, and ATFL in enhancing accuracy, robustness, and efficiency. Code and dataset are publicly available at https://github.com/chen-yuzhi/YOLO11-AU-IR .
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Hebei University of Architecture, Zhangjiakou, China.
- Hebei Digital Education Collaborative Innovation, Shijiazhuang, China.
- College of Computer and Cyber Security, Hebei Normal University, Shijiazhuang, China.
ORCID-Profile: YuZhi Chen, HaoYue Sun
Förderung
- Doctoral (Postdoctoral) Fund of Hebei Normal University
- Hebei Provincial Department of Science and Technology
Zitieren
APA
Chen, Y., Sun, H., Tian, L., Yang, Y., Wang, S., & Wang, T. (2025). Detecting infrared UAVs on edge devices through lightweight instance segmentation. PLOS One, 20(8), e0330074. https://doi.org/10.1371/journal.pone.0330074
Vancouver
Chen Y, Sun H, Tian L, Yang Y, Wang S, Wang T. Detecting infrared UAVs on edge devices through lightweight instance segmentation. PLOS One. 2025;20(8):e0330074. doi:10.1371/journal.pone.0330074
BibTeX
@article{Chen2025,
author = {Chen, YuZhi and Sun, HaoYue and Tian, Liang and Yang, Ye and Wang, ShenYang and Wang, TianYou},
title = {Detecting infrared UAVs on edge devices through lightweight instance segmentation},
journal = {PLOS One},
year = {2025},
volume = {20},
number = {8},
pages = {e0330074},
doi = {10.1371/journal.pone.0330074}
}
EUVIMED