Video Spatial Grounding on ST-Align
60.3vIoU@0.3STVG-R1
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| STVG-R12026.02 | 60.3 | 53.9 | 48.6 | |
| Qwen2.5-VL-7BVisual Prompting=true2026.02 | 57.7 | 51.4 | 46.6 | |
| LLava-ST-7B2026.02 | 47.2 | 30.9 | 32.5 | |
| Qwen2.5-VL-7BVisual Prompting=false2026.02 | 44.6 | 39.5 | 35.5 | |
| GroundingGPT-7B2026.02 | 19.7 | 5.4 | 17.9 |