Object Detection on MAVREC aerial (val)
65.35mAPFlorence-2 + CPA + PCL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Florence-2 + CPA + PCLBackbone=Florence-2-base (230M parameters)2026.05 | 65.35 | 76.79 | 70.16 | 62.74 | 57.41 | |
| Florence-2 + Complexity-Aware Pathway AggregationBackbone=Florence-2-base (230M parameters)2026.05 | 64.49 | 76 | 69.58 | 61.8 | 59.66 | |
| Florence-2 + Paired Curriculum LearningBackbone=Florence-2-base (230M parameters)2026.05 | 64.37 | 76.62 | 69.63 | 63.77 | 51.92 | |
| Florence-2 (random)Backbone=Florence-2-base (230M parameters)2026.05 | 63.73 | 75.22 | 68.62 | 62.08 | 49.26 | |
| YOLOv72026.05 | 31.3 | 57.7 | — | 34.2 | 61.2 | |
| D-DETRPre-trained on COCO=true2026.05 | 31 | 61.7 | — | 31.7 | 55.1 | |
| YOLO-NAS-L2026.05 | 30.3 | 49.8 | — | 29.2 | 61.5 | |
| DETR2026.05 | 24.9 | 39.7 | — | 27.6 | 45.3 | |
| D-DETRPre-trained on COCO=false2026.05 | 13.1 | 28.3 | — | 14.2 | 38.1 |