Visual Grounding on D3 Intra-scenario
37.5APb (Full)APE (D)
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| APE (D)Backbone=ViT-L2023.12 | 37.5 | 34.4 | 38.8 | 35.4 | 33.9 | 31.7 | |
| APE (B)Backbone=ViT-L2023.12 | 30 | 26.8 | 29.9 | 26.6 | 30.3 | 27.3 | |
| APE (C)Backbone=ViT-L2023.12 | 27.8 | 25.6 | 27.9 | 25.5 | 27.3 | 25.8 | |
| APE (A)Backbone=ViT-L2023.12 | 25.1 | 22.6 | 24.5 | 22 | 26.9 | 24.4 | |
| OFA-DODBackbone=R1012023.12 | 21.6 | — | 23.7 | — | 15.4 | — | |
| G-DINOBackbone=Swin-B2023.12 | 20.7 | — | 20.1 | — | 22.5 | — | |
| UNINEXTBackbone=ViT-H2023.12 | 20 | — | 20.6 | — | 18.1 | — | |
| OWL-ViTBackbone=ViT-L2023.12 | 9.6 | — | 10.7 | — | 6.4 | — | |
| CORABackbone=R502023.12 | 6.2 | — | 6.7 | — | 5 | — | |
| OFABackbone=R1522023.12 | 4.2 | — | 4.1 | — | 4.6 | — |