Video Temporal Grounding on Charades-STA original (test)
60.8mIoUVideoChat2-7B
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| VideoChat2-7Bsource=original paper2025.12 | 60.8 | — | 71.7 | 50.2 | |
| TimeChat-7Bsource=original paper2025.12 | 58.1 | 78.1 | 60.8 | 35.3 | |
| MiMo-VL-7Bsource=original paper2025.12 | 50 | — | — | — | |
| VideoChat-Flash-7Bsource=original paper2025.12 | 45.2 | 72.5 | 51.4 | 26.4 | |
| Qwen2.5-VL-7BmIoU_source=original paper, Status=Baseline2025.12 | 43.6 | 59.4 | 38.2 | 18.1 | |
| TimeLens-7BModel Category=Open-Source2025.12 | 42.3 | 70.7 | 39.8 | 14.5 | |
| Gemini-1.0-ProModel Category=Proprietary2025.12 | 35.1 | 55.6 | 29 | 9.5 | |
| GPT-4oModel Category=Proprietary2025.12 | 34.7 | 51.6 | 27.9 | 11.7 | |
| Gemini-1.5-ProModel Category=Proprietary2025.12 | 34.6 | 53.9 | 25.5 | 8.8 | |
| Gemini-1.5-FlashModel Category=Proprietary2025.12 | 31.1 | 47 | 21.8 | 7.1 | |
| GPT-4VModel Category=Proprietary2025.12 | 28.4 | 39.7 | 18.3 | 6.2 |