Open-vocabulary Object Detection on OV-COCO (novel)
44.3AP@50 (Novel)F-ViT+CLIPSelf
Evaluation Results
| Method | Links | |
|---|---|---|
| F-ViT+CLIPSelfBackbone=ViT-L/142023.10 | 44.3 | |
| CORA+Backbone=RN50x42023.10 | 43.1 | |
| CORABackbone=RN50x42023.10 | 41.7 | |
| F-ViT+CLIPSelfBackbone=ViT-B/162023.10 | 37.6 | |
| CFM-ViTBackbone=ViT-L/162023.10 | 34.1 | |
| BARON-KDBackbone=RN502023.10 | 34 | |
| RO-ViTBackbone=ViT-L/162023.10 | 33 | |
| PromptOVDBackbone=ViT-B/162023.10 | 30.6 | |
| OV-DETRBackbone=RN502023.10 | 29.4 | |
| F-VLMBackbone=RN502023.10 | 28 | |
| DeticBackbone=RN502023.10 | 27.8 | |
| ViLDBackbone=RN502023.10 | 27.6 | |
| F-ViTBackbone=ViT-L/142023.10 | 24.7 | |
| F-ViTBackbone=ViT-B/162023.10 | 17.5 |