Multi-segment grounding on Perception Test (val)
31.7F1 ScoreMUSEG-7B
Evaluation results
| Method | Links | |
|---|---|---|
| MUSEG-7BCategory=Open-source ~ 7B Models2025.05 | 31.7 | |
| Qwen2.5-VL-7B + vanilla GRPOCategory=Open-source ~ 7B Models, Training=vanilla GRPO2025.05 | 30 | |
| MUSEG-3BCategory=Open-source ~ 3B Models2025.05 | 29.1 | |
| VideoChat-R1Category=Open-source ~ 7B Models2025.05 | 27.1 | |
| TimeZeroCategory=Open-source ~ 7B Models2025.05 | 26.8 | |
| Qwen2.5-VL-7BCategory=Open-source ~ 7B Models2025.05 | 25.3 | |
| TEMPURACategory=Open-source ~ 3B Models2025.05 | 20.7 | |
| Qwen2.5-VL-7B + vanilla SFTCategory=Open-source ~ 7B Models, Training=vanilla SFT2025.05 | 20.3 | |
| Qwen2.5-VL-3BCategory=Open-source ~ 3B Models2025.05 | 19.4 | |
| TRACE-7BCategory=Open-source ~ 7B Models2025.05 | 14 | |
| E.T. ChatCategory=Open-source ~ 7B Models2025.05 | 9.2 | |
| Video-R1Category=Open-source ~ 7B Models2025.05 | 5.7 |