Object Detection on Golden (test)
86.8mAP50Weakly Supervised School Detection Framework (YOLO26n)
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Weakly Supervised School Detection Framework (YOLO26n)Training Strategy=Auto-labeled pre-training + Golden fine-tuning, Training Samples=1002026.05 | 86.8 | 78.2 | 85.5 | 81.7 | 67.4 | |
| Weakly Supervised School Detection Framework (Faster R-CNN)Training Strategy=Auto-labeled pre-training + Golden fine-tuning, Training Samples=1002026.05 | 71.2 | 63.1 | 71.4 | 67 | 48.1 | |
| Faster R-CNNTraining Strategy=Auto-labeled, Training Samples=1002026.05 | 60 | 19.3 | 86.9 | 31.6 | 39.5 | |
| Faster R-CNNTraining Strategy=Golden + ECP, Training Samples=1002026.05 | 58.7 | 52.2 | 71.4 | 60.3 | 33.7 | |
| YOLO26nTraining Strategy=Auto-labeled, Training Samples=1002026.05 | 46.4 | 47.8 | 51.2 | 49.4 | 33.2 | |
| YOLO26nTraining Strategy=Golden + ECP, Training Samples=1002026.05 | 32.2 | 28.1 | 42.9 | 34 | 15.9 |