Visual Grounding on MM-Conv Partitive NP 1.0 (full)
29.6M@0.5Qwen2.5-VL
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5-VLContext=w/ ctx2026.05 | 29.6 | 34.2 | — | |
| Qwen2.5-VLContext=w/o ctx2026.05 | 22.5 | 26.4 | — | |
| Florence-2Context=w/ ctx2026.05 | 21.6 | 23.6 | — | |
| Kosmos-2Context=w/ ctx2026.05 | 19.7 | 22.2 | — | |
| FerretContext=w/ ctx2026.05 | 18.3 | 21.6 | — | |
| Kosmos-2Context=w/o ctx2026.05 | 16.5 | 17.9 | — | |
| Florence-2Context=w/o ctx2026.05 | 16.3 | 18.8 | — | |
| GroundingGPTContext=w/ ctx2026.05 | 14.7 | 20.9 | — | |
| FerretContext=w/o ctx2026.05 | 14.7 | 16.7 | — | |
| GroundingGPTContext=w/o ctx2026.05 | 14.2 | 16.3 | — |