EUVIMEDEUVIMED entdecken
Crossref

DS-DETR: Dynamic Sparse Detection Transformer for Small Object Detection in High-Resolution Remote Sensing Images

Bo Peng, Yifan Chen, Chao Ma, Bin Jiang, Ningsheng Liao, Mi Zhu

IEEE Geoscience and Remote Sensing Letters · 2026 · 23 · 1-5

DOI: 10.1109/lgrs.2026.3673202 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Geoscience and Remote Sensing Letters, 2026. Verfasst von Bo Peng und 5 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben. Gefördert von: Chongqing Natural Science Foundation Ph.D. “Direct Train” Project, Chongqing Technology Innovation and Application Development Project, Major Synchronous Implementation Project of Chongqing University of Technology.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Remote sensing images have been widely applied in both military and civilian fields. However, targets in high-resolution remote sensing images are typically extremely small, with features easily intertwined with complex backgrounds, leading to feature aliasing and poor discriminability. Additionally, the high computational cost of traditional attention mechanisms limits the practical deployment of transformer-based detectors. To address these challenges, we propose dynamic sparse-detection transformer (DS-DETR), a detection transformer that integrates dynamic feature enhancement and sparse attention mechanisms, developed based on real-time (RT)-DETR. It incorporates three key innovations: first, the dynamic feature enhancement network (DFENet) reduces the parameters of the baseline while enhancing the features of weak and small targets; second, the adaptive sparse attention (ASA) mechanism focuses on potential target regions and suppresses redundant background features, while reducing computational complexity; finally, the detail refinement module refines the local textures of targets. The experimental results on the DOTA v1.5 dataset show that the proposed model improves mAP by 2.5, reduces parameters by 5.6M, and increases FPS from 94 to 112 compared to the baseline. It also exhibits stable performance gains on the NWPU VHR-10 dataset. These results indicate that DS-DETR provides a reliable solution for RT and high-precision remote sensing object detection. The code will be available at https://github.com/yfchenbruce/DS-DETR

Förderung

  • Chongqing Natural Science Foundation Ph.D. “Direct Train” Project
  • Chongqing Technology Innovation and Application Development Project
  • Major Synchronous Implementation Project of Chongqing University of Technology

Zitieren

APA

Peng, B., Chen, Y., Ma, C., Jiang, B., Liao, N., & Zhu, M. (2026). DS-DETR: Dynamic Sparse Detection Transformer for Small Object Detection in High-Resolution Remote Sensing Images. IEEE Geoscience and Remote Sensing Letters, 23, 1-5. https://doi.org/10.1109/lgrs.2026.3673202

Vancouver

Peng B, Chen Y, Ma C, Jiang B, Liao N, Zhu M. DS-DETR: Dynamic Sparse Detection Transformer for Small Object Detection in High-Resolution Remote Sensing Images. IEEE Geoscience and Remote Sensing Letters. 2026;23:1-5. doi:10.1109/lgrs.2026.3673202

BibTeX
@article{Peng2026,
  author = {Peng, Bo and Chen, Yifan and Ma, Chao and Jiang, Bin and Liao, Ningsheng and Zhu, Mi},
  title = {DS-DETR: Dynamic Sparse Detection Transformer for Small Object Detection in High-Resolution Remote Sensing Images},
  journal = {IEEE Geoscience and Remote Sensing Letters},
  year = {2026},
  volume = {23},
  pages = {1--5},
  doi = {10.1109/lgrs.2026.3673202}
}
Kostenlos starten →