Research Article
WS-YOLO: A single-stage detector designed for aircraft detection by enhancing fine-grained segmentation
@INPROCEEDINGS{10.4108/eai.18-12-2025.2365272, author={Yuzhu Lei and Jun Li and Lei Zhang and Guoming Song}, title={WS-YOLO: A single-stage detector designed for aircraft detection by enhancing fine-grained segmentation}, proceedings={Proceedings of the 13th International Conference on Identification, Information and Knowledge in the Internet of Things, IIKI 2025, 18-21 December 2025, Chengdu, China}, publisher={EAI}, proceedings_a={IIKI}, year={2026}, month={6}, keywords={Aircraft Detection Remote Sensing Images YOLO}, doi={10.4108/eai.18-12-2025.2365272} }- Yuzhu Lei
Jun Li
Lei Zhang
Guoming Song
Year: 2026
WS-YOLO: A single-stage detector designed for aircraft detection by enhancing fine-grained segmentation
IIKI
EAI
DOI: 10.4108/eai.18-12-2025.2365272
Abstract
With the advancement of aerospace technology, small object detection in satellite remote sensing has become a major research focus globally, with aircraft detection regarded as a fundamental task. Advances in sensor technology enable the capture of more detailed aircraft features, yet simultaneously make fine-grained detection increasingly challenging. To address this, this paper proposes WS-YOLO, a novel architecture incorporating the Enhanced Fine-grained Module. Experimental results on the CORS-ADD and MAR20 datasets show that WS-YOLO achieves mAP improvements of 0.44%–4.95% and 0.44%–1.78%, respectively, compared to six existing YOLO variants, demonstrating its effectiveness in detecting small aircraft targets under high fine-grained conditions.
