Moment Retrieval on DiDeMo (test)
46.3R@1 (IoU=0.3)GroundVTS-Q
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GroundVTS-QBackbone=Qwen2.5-VL-7B2026.04 | 46.3 | 27.8 | 30 | |
| VTimeLLM2026.04 | 45 | 28.8 | 27.9 | |
| HawkEye2026.04 | 44.8 | 29.7 | 29.5 | |
| TimeChat2026.04 | 42.8 | 24.4 | 28.2 | |
| Momenter2026.04 | 38.2 | 21.8 | 26.5 | |
| GroundVTS-IBackbone=InternVL3.5-8B2026.04 | 37.5 | 22.3 | 24.2 | |
| InternVL-G2026.04 | 36.6 | 21 | 23.1 | |
| VideoChat2026.04 | 34.5 | 14.5 | 22.4 | |
| Valley2026.04 | 33.2 | 13.4 | 21.8 | |
| QwenVL-G2026.04 | 32.6 | 17.7 | 22 | |
| InternVL3.5-8B2026.04 | 29.5 | 23.9 | 23 | |
| Qwen2.5-VL-7B2026.04 | 28.7 | 22.7 | 22.2 | |
| Video-LLaMA2026.04 | 20.1 | 8.2 | 14.3 | |
| Video-ChatGPT2026.04 | 19.8 | 6.5 | 13.7 |