MSRP-TODNet: a multi-scale reinforced region wise analyser for tiny object detection
Experimental results on the DOTA dataset demonstrate that MSRP-TODNet outperforms existing state-of-the-art methods. Specifically, it achieves an mAP @0.5 of 84.2%, mAP @0.5:0.95 of 54.1%, and an F1-Score of 84.0%, surpassing improved TPH-YOLOv5, YOLOv7-Tiny, and DRDet by margins of 1.7%-6.1% in detection performance. These results demonstrate the framework's effectiveness for accurate, real-time small object detection in UAV surveillance and aerial imagery.
Authors
- K P N V Satya Sree (ORCID: https://orcid.org/0000-0003-4567-3447)
- Thulasi Bikku (ORCID: https://orcid.org/0000-0001-6202-1438)
- Srinivasarao Thota (ORCID: https://orcid.org/0000-0002-3265-5656)
- P. Shanmugasundaram (ORCID: https://orcid.org/0000-0003-1176-0411)
- Malligunta Kiran Kumar
Institutions
- Rama University (IN)
- Amrita Vishwa Vidyapeetham (IN)
- Koneru Lakshmaiah Education Foundation (IN)
Publication Details
- Journal
- BMC Research Notes
- Published
- 2025-04-30
- DOI
- https://doi.org/10.1186/s13104-025-07263-7
- Citations
- 4
- Primary Topic
- Advanced Neural Network Applications
- Type
- article
- Field-Weighted Citation Impact
- 5.10