Video Grounding on E.T. Bench-Grounding (full subset)
35.5Mean F1Tarsier2-7B
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Tarsier2-7BModel Type=Open-source, Parameter Scale=<10B, Fine-tuned on E.T. Instruct 164K=true2025.01 | 35.5 | 38.4 | 11 | 31.8 | 19.4 | 66.8 | |
| E.T. ChatModel Type=Open-source, Parameter Scale=<10B, Fine-tuned on E.T. Instruct 164K=true2025.01 | 33.5 | 38.6 | 10.2 | 30.8 | 25.4 | 62.5 | |
| Qwen2-VL-7BModel Type=Open-source, Parameter Scale=<10B, Fine-tuned on E.T. Instruct 164K=true2025.01 | 33.5 | 39.7 | 7 | 26.9 | 17.1 | 66.9 | |
| Tarsier-7BModel Type=Open-source, Parameter Scale=<10B, Fine-tuned on E.T. Instruct 164K=true2025.01 | 30.9 | 39.6 | 9 | 25 | 25.4 | 47.6 | |
| Gemini-1.5-FlashModel Type=Proprietary, Tested on subset=true2025.01 | 28.5 | 43.9 | 5.4 | 27 | 5.4 | 60.8 | |
| GPT-4oModel Type=Proprietary, Tested on subset=true2025.01 | 27.9 | 40.4 | 4.5 | 20 | 17.6 | 56.9 | |
| Gemini-1.5-ProModel Type=Proprietary, Tested on subset=true2025.01 | 27.6 | 43.1 | 6.2 | 33.8 | 7.9 | 47 | |
| GPT-4VModel Type=Proprietary, Tested on subset=true2025.01 | 26.1 | 27 | 1.8 | 18 | 28.6 | 55.1 | |
| TimeChatModel Type=Open-source, Parameter Scale=<10B, Fine-tuned on E.T. Instruct 164K=true, Tested on subset=true2025.01 | 24.3 | — | — | — | — | — | |
| VTG-LLMModel Type=Open-source, Parameter Scale=<10B, Tested on subset=true2025.01 | 21.8 | 15.9 | 3.7 | 14.4 | 26.8 | 48.2 | |
| LITAModel Type=Open-source, Parameter Scale=<10B, Tested on subset=true2025.01 | 19.7 | 22.2 | 4.6 | 18 | 29.7 | 23.9 |