Multimodal Cloze on Omni-Cloze
57Visual ScoreOmni-Captioner-7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Omni-Captioner-7BModel Type=Open-Source Models2025.10 | 57 | 54.5 | 62.1 | 56.4 | |
| Omni-Captioner-7BModel Category=Open-Source Models2025.10 | 57 | 54.5 | 62.1 | 56.4 | |
| Gemini 2.5 ProModel Type=Proprietary Models2025.10 | 40.8 | 44.1 | 52.8 | 43.6 | |
| Gemini 2.5 ProModel Category=Proprietary Models2025.10 | 40.8 | 44.1 | 52.8 | 43.6 | |
| video-SALMONN 2-7BModel Type=Open-Source Models2025.10 | 37.5 | 40.3 | 45 | 39.5 | |
| video-SALMONN 2-7BModel Category=Open-Source Models2025.10 | 37.5 | 40.3 | 45 | 39.5 | |
| Gemini 2.0 FlashModel Type=Proprietary Models2025.10 | 32.3 | 31.7 | 40.1 | 33.2 | |
| Gemini 2.0 FlashModel Category=Proprietary Models2025.10 | 32.3 | 31.7 | 40.1 | 33.2 | |
| Gemini 2.5 FlashModel Type=Proprietary Models2025.10 | 24.4 | 36.2 | 41.4 | 33 | |
| Gemini 2.5 FlashModel Category=Proprietary Models2025.10 | 24.4 | 36.2 | 41.4 | 33 | |
| Qwen2.5-Omni-7BModel Type=Open-Source Models2025.10 | 18.3 | 14.1 | 21.9 | 16.6 | |
| Qwen2.5-Omni-7BModel Category=Open-Source Models2025.10 | 18.3 | 14.1 | 21.9 | 16.6 | |
| VideoLLaMA 2-7BModel Type=Open-Source Models2025.10 | 7.6 | 4.3 | 8.7 | 6.6 | |
| VideoLLaMA 2-7BModel Category=Open-Source Models2025.10 | 7.6 | 4.3 | 8.7 | 6.6 | |
| video-SALMONN-13BModel Type=Open-Source Models2025.10 | 3.5 | 2.3 | 4.6 | 3.3 | |
| video-SALMONN-13BModel Category=Open-Source Models2025.10 | 3.5 | 2.3 | 4.6 | 3.3 |