Object Detection on WildDash (train)
34.7mAP50Unified detector (UniDet)
Evaluation results
| Method | Links | |
|---|---|---|
| Unified detector (UniDet)Strategy=Unified (retrained), Backbone=ResNet-502021.02 | 34.7 | |
| Partitioned DetectorStrategy=Partitioned, Backbone=ResNet-502021.02 | 34.4 | |
| EnsembleStrategy=Ensemble, Backbone=ResNet-502021.02 | 32.1 | |
| Objects365 modelTraining Data=Objects365, Backbone=ResNet-502021.02 | 27.6 | |
| COCO modelTraining Data=COCO, Backbone=ResNet-502021.02 | 25.9 | |
| Mapillary modelTraining Data=Mapillary, Backbone=ResNet-502021.02 | 23.4 | |
| OpenImages modelTraining Data=OpenImages, Backbone=ResNet-502021.02 | 20.3 |