Object Detection On Nao
Métriques
mAP
mAP w/o OOD
mAR
Résultats
Résultats de performance de divers modèles sur ce benchmark
Nom du modèle | mAP | mAP w/o OOD | mAR | Paper Title | Repository |
|---|---|---|---|---|---|
| EfficientDet-D2 | 12.8 | 25.4 | 40.2 | Natural Adversarial Objects | - |
| EfficientDet-D4 | 15.0 | 29.6 | 42.7 | Natural Adversarial Objects | - |
| EfficientDet-D7 | 13.6 | 26.6 | 40.8 | Natural Adversarial Objects | - |
| Faster RCNN | 13.5 | 22.8 | 41.4 | Natural Adversarial Objects | - |
| RetinaNet-R50 | 11.1 | 19.5 | 37.2 | Natural Adversarial Objects | - |
| YOLOv3 | 10.0 | 17.5 | 28.4 | Natural Adversarial Objects | - |
| Mask RCNN R50 | 15.2 | 24.6 | 43.8 | Natural Adversarial Objects | - |
0 of 7 row(s) selected.