Video Referring on VideoRefer-Bench-D
4.92SCSWIM
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SWIMPrompt types=Text2026.05 | 4.92 | 3.85 | 3.43 | 2.96 | 3.78 | |
| TraceVision-7BParameters=7B2026.02 | 4.73 | 3.89 | 3.11 | 3.28 | 3.83 | |
| DAM-8BParameters=8B2026.02 | 4.69 | 3.61 | 3.34 | 3.09 | 3.68 | |
| DAM-8BPrompt types=Point, Box, Mask2026.05 | 4.69 | 3.61 | 3.34 | 3.03 | 3.68 | |
| VideoRefer-7BParameters=7B2026.02 | 4.44 | 3.27 | 3.1 | 3.04 | 3.46 | |
| VideoRefer-7BPrompt types=Mask2026.05 | 4.44 | 3.27 | 3.1 | 3.04 | 3.46 | |
| Qwen2.5-VL-7BPrompt types=Text2026.05 | 4.33 | 3.19 | 2.88 | 2.58 | 3.24 | |
| GPT-4oPrompt types=Text2026.05 | 4.15 | 3.31 | 3.11 | 2.43 | 3.25 | |
| InternVL2-26BPrompt types=Text2026.05 | 4.08 | 3.35 | 3.08 | 2.28 | 3.2 | |
| Qwen2-VL-7BPrompt types=Text2026.05 | 3.99 | 3.05 | 2.44 | 2.44 | 2.97 | |
| PAM-3BParameters=3B2026.02 | 3.92 | 2.84 | 2.88 | 2.94 | 3.14 | |
| Qwen2.5-Omni-7BPrompt types=Text2026.05 | 3.92 | 2.82 | 1.97 | 2.34 | 2.76 | |
| PAM-3BPrompt types=Point, Box, Mask2026.05 | 3.92 | 2.84 | 2.88 | 2.94 | 3.14 | |
| GPT-4o-miniPrompt types=Text2026.05 | 3.89 | 3.18 | 2.62 | 2.5 | 3.05 | |
| Artemis-7BParameters=7B2026.02 | 3.42 | 1.34 | 1.39 | 2.9 | 2.26 | |
| Artemis-7BPrompt types=Box2026.05 | 3.42 | 1.34 | 1.39 | 2.9 | 2.26 | |
| Osprey-7BPrompt types=Point, Mask2026.05 | 3.3 | 2.66 | 2.1 | 1.58 | 2.41 | |
| Ferret-7BPrompt types=Point, Box, Mask2026.05 | 3.2 | 2.38 | 1.97 | 1.38 | 2.23 | |
| LLaVA-OV-7BPrompt types=Text2026.05 | 3.09 | 1.94 | 2.5 | 2.41 | 2.48 | |
| LongVA-7BPrompt types=Text2026.05 | 3.02 | 2.3 | 1.92 | 2.51 | 2.44 | |
| Elysium-7BParameters=7B2026.02 | 2.35 | 0.3 | 0.02 | 3.59 | 1.57 | |
| Elysium-7BPrompt types=Box2026.05 | 2.35 | 0.3 | 0.02 | 3.59 | 1.57 | |
| LongVU-7BPrompt types=Text2026.05 | 2.33 | 1.8 | 2.39 | 1.68 | 2.05 |