Table 8 Comparison results on the VEDAI database.
From: Multi path attention and scale aware fusion for accurate object detection in remote sensing imagery
Methods | Epoch | #Params | GFLOPs | FPS | \({\text {mAP}}_{50}\) | \({\text {mAP}}_{50:95}\) |
|---|---|---|---|---|---|---|
YOLOv8N | 300 | 5.5 | 6.5 | 159.1 | 52.8 | 33.1 |
YOLOv10N | 300 | 5.5 | 6.5 | 183.3 | 47.9 | 30.5 |
YOLOv11N | 300 | 5.3 | 6.3 | 163.4 | 50.4 | 30.8 |
YOLOv12N | 300 | 5.2 | 5.8 | 95.6 | 48.4 | 31.9 |
YOLOv13N | 300 | 5.2 | 6.1 | 67.5 | 45.3 | 27.9 |
RT-DETR-r18 | 160 | 38.6 | 57.0 | 29.8 | 52.8 | 34.2 |
RT-DETRv2-r18 | 160 | 19.9 | 59.921 | 201.36 | 50.8 | 34.2 |
D-Fine-N | 160 | 14.5 | 7.1267 | 204.95 | 38.4 | 23.8 |
DEIM-N | 160 | 14.5 | 7.1267 | 176.50 | 44.8 | 28.1 |
HyperFusion-DEIM-N | 160 | 134.1 | 79.7097 | 200.64 | 53.3 | 34.6 |
