LeanCOD: Real-Time Small Camouflaged Object Detection on Edge Devices
Sensors · 09.07.2026 · 26 (14) · 4354
DOI: 10.3390/s26144354 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 09.07.2026. Verfasst von Youngjin Kim und 2 weiteren Autor:innen. Beteiligte Einrichtung: Research and Development Center, dSPECTER, Seongnam 13449, Republic of Korea. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of SMEs and Startups.
Abstract (Originalsprache)
Camouflaged object detection (COD) methods suffer severe degradation on small camouflaged objects and remain incapable of real-time inference on edge devices. We propose LeanCOD, a framework that pairs a strong foundation-model encoder with a lightweight decoder to enable high-resolution inference. A size-aware composite loss further strengthens supervision on small camouflaged objects. Our size-wise experiments reveal that the 0–1% extra-small-object regime is the major performance bottleneck for existing COD methods. LeanCOD achieves an Sα of 0.915 on COD10K at a 576×576 resolution, outperforming competing methods at equal or lower resolutions. Deployed with TensorRT FP16 on an NVIDIA Jetson AGX Orin, LeanCOD runs at 31.6 FPS while maintaining an Sα of 0.908 at a 576×576 resolution, exceeding the 30 FPS real-time threshold.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Research and Development Center, dSPECTER, Seongnam 13449, Republic of Korea.
ORCID-Profile: Youngjin Kim, Dong He, Young Hoo Cho
Förderung
- Ministry of SMEs and Startups
Zitieren
APA
Kim, Y., He, D., & Cho, Y. H. (2026). LeanCOD: Real-Time Small Camouflaged Object Detection on Edge Devices. Sensors, 26(14), 4354. https://doi.org/10.3390/s26144354
Vancouver
Kim Y, He D, Cho YH. LeanCOD: Real-Time Small Camouflaged Object Detection on Edge Devices. Sensors. 2026;26(14):4354. doi:10.3390/s26144354
BibTeX
@article{Kim2026,
author = {Kim, Youngjin and He, Dong and Cho, Young Hoo},
title = {LeanCOD: Real-Time Small Camouflaged Object Detection on Edge Devices},
journal = {Sensors},
year = {2026},
volume = {26},
number = {14},
pages = {4354},
doi = {10.3390/s26144354}
}
EUVIMED