Visual Grounding on MS-COCO 2014
43.2mIoUGROUNDIT
Evaluation Results
| Method | Links | |
|---|---|---|
| GROUNDITBackbone=PixArt-α [8]2024.10 | 43.2 | |
| PixArt-R&BBackbone=PixArt-α [8]2024.10 | 41.8 | |
| R&BBackbone=Stable Diffusion [41]2024.10 | 41.1 | |
| BoxDiffBackbone=Stable Diffusion [41]2024.10 | 32.4 | |
| Layout-GuidanceBackbone=Stable Diffusion [41]2024.10 | 30.7 | |
| Attention-RefocusingBackbone=Stable Diffusion [41]2024.10 | 25.4 | |
| PixArt-αBackbone=Stable Diffusion [41]2024.10 | 23.3 | |
| Stable DiffusionBackbone=Stable Diffusion [41]2024.10 | 17.6 |