Table 3 Results of our model compared to existing excellent methods on the UAVDT dataset. The best results are highlighted in bold.
Method | P | R | Cars | Truck | Bus | Para | FPS | mAP |
|---|---|---|---|---|---|---|---|---|
FEYOLO62 | 96.3 | 95.0 | 73.2 | 83.5 | 84.2 | 6.3M | 155 | 80.3 |
AD-YOLO57 | 97.1 | 95.4 | 74.0 | 85.5 | 91.1 | 7.1M | 135 | 83.5 |
MobileNetv3-SSD64 | 92.1 | 85.4 | 50.5 | 67.6 | 68.8 | 3.93M | 152 | 62.4 |
YOLOv7tiny-silu65 | 95.3 | 94.8 | 61.4 | 76.3 | 78.6 | 6.1M | 233 | 72.1 |
YOLOv7tiny-mobilevit-odconv66 | 94.6 | 88.9 | 52.3 | 69.8 | 72.2 | 8.6M | 182 | 64.8 |
YOLOv8n32 | 96.1 | 95.2 | 70.3 | 82.5 | 83.1 | 3.2M | 248 | 78.7 |
YOLOv8s32 | 96.7 | 96.5 | 76.2 | 88.9 | 89.3 | 11.2M | 218 | 84.7 |
YOLO11n33 | 96.2 | 95.2 | 71.7 | 82.3 | 83.2 | 2.6M | 281 | 79.0 |
YOLO11s33 | 97.8 | 97.9 | 77.2 | 88.6 | 89.7 | 9.4M | 229 | 85.1 |
YOLOv12n56 | 95.8 | 94.8 | 69.0 | 81.0 | 81.3 | 2.5M | 254 | 77.1 |
YOLOv12s56 | 96.5 | 95.9 | 76.2 | 87.6 | 88.8 | 9.1M | 211 | 84.2 |
LUDY-n27 | 97.5 | 97.2 | 74.3 | 85.2 | 86.9 | 2.8M | 247 | 82.1 |
LUDY-s27 | 98.2 | 98.3 | 80.7 | 91.8 | 92.6 | 10.4M | 211 | 88.4 |
YOLO-UD-n | 97.9 | 98.0 | 77.2 | 89.3 | 90.2 | 3.3M | 258 | 85.6 |
YOLO-UD-s | 98.6 | 98.4 | 81.3 | 94.9 | 95.6 | 12.1M | 216 | 90.6 |
