Spatial Grounding on RefCOCO (val)
90.6AccuracyG-DINO-L
Evaluation Results
| Method | Links | |
|---|---|---|
| G-DINO-L2024.12 | 90.6 | |
| Shikra-7B2024.12 | 87 | |
| VisionLLM-HRefined decoder=true2024.12 | 86.7 | |
| VideoChat-TPO2024.12 | 85.9 | |
| NEXT-Chat-7B2024.12 | 85.5 | |
| OFA-L2024.12 | 80 | |
| DDPNBackbone=ResNet-1012018.05 | 76.8 | |
| MAttNet2024.12 | 76.4 | |
| DDPNBackbone=VGG-162018.05 | 73.4 | |
| Yu et al.2018.05 | 69.5 | |
| Li et al.2018.05 | 52.4 | |
| Wu et al.2018.05 | 48.2 | |
| SCRC2018.05 | 19 |