YOLOv11 Model Tested on Drone Military Detection Across Visual Spectrums
A new study evaluates military object detection from drone imagery across multiple visual spectrums using the YOLOv11-small model. Researchers created four datasets—Gray Scale, Thermal Vision, Night Vision, and Obscura Vision—to simulate real-world conditions like low visibility, heat-based imagery, and nighttime. The base dataset is KIIT-MiTA, which contains drone images of military scenarios but lacks diverse environmental conditions. The study aims to assess model performance under varying operational environments. The research is published on arXiv under identifier 2605.21157.
Key facts
- The study uses YOLOv11-small model for military object detection.
- Four datasets simulate real-world conditions: Gray Scale, Thermal Vision, Night Vision, Obscura Vision.
- Base dataset is KIIT-MiTA, comprising drone images of military scenarios.
- Research published on arXiv with identifier 2605.21157.
- Datasets address low visibility, heat-based imagery, and nighttime conditions.
- Drones are essential for intelligence and precise attacks in modern warfare.
- KIIT-MiTA does not account for various real-world scenarios.
- The study evaluates model performance under varying conditions.
Entities
Institutions
- KIIT-MiTA
- arXiv