Visual Grounding on Talk2Car
80.12IoUE3AD
Evaluation Results
| Method | Links | |
|---|---|---|
| E3ADBackbone=VLM2025.12 | 80.12 | |
| CAVGBackbone=ViT2025.12 | 74.62 | |
| RSD-LXMERTBackbone=ResNet-1012025.12 | 72.64 | |
| UNINEXTBackbone=ResNet-502025.12 | 70.87 | |
| MDERTBackbone=ResNet-1012025.12 | 70.52 | |
| VL-BERTBackbone=ResNet-1012025.12 | 70.03 | |
| CMRTBackbone=ResNet-1522025.12 | 69.11 | |
| CMSVGBackbone=EfficientNet2025.12 | 68.61 | |
| Grounding-DINOBackbone=ViT2025.12 | 68.15 | |
| TransVGBackbone=ResNet-1012025.12 | 65.83 | |
| VLTVGBackbone=ResNet-1012025.12 | 63.33 | |
| AttnGrounderBackbone=ResNet-502025.12 | 61.32 | |
| Qwen3-VL-8BBackbone=VLM2025.12 | 56.19 | |
| Qwen2.5-VL-72BBackbone=VLM2025.12 | 56.17 | |
| Qwen2.5-VL-7BBackbone=VLM2025.12 | 47.31 |