Multi-image evaluation on MI-Bench
86.4GCmPLUG-Owl3
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| mPLUG-Owl3Model Type=Open-source2024.08 | 86.4 | 70.1 | 33 | 46.8 | 67.2 | 50.1 | 31.1 | 48.8 | |
| Idefics2-IModel Type=Open-source2024.08 | 83.1 | 49.7 | 32.6 | 44.8 | 56.4 | 43.9 | 25.6 | 39 | |
| MantisModel Type=Open-source2024.08 | 83 | 54.1 | 37.6 | 45.5 | 63.4 | 37.7 | 26.4 | 41.7 | |
| GPT-4oModel Type=Closed-source2024.08 | 80.7 | 90.5 | 46.8 | 68 | 69.8 | 74.8 | 54.7 | 63.3 | |
| GPT-4VModel Type=Closed-source2024.08 | 72.8 | 79.2 | 45.8 | 61.8 | 66.3 | 71 | 52 | 56 | |
| mPLUG-Owl2Model Type=Open-source2024.08 | 64.2 | 40.1 | 35.6 | 30.7 | 41.3 | 39 | 17 | 25.6 | |
| MMICLModel Type=Open-source2024.08 | 53.7 | 46.4 | 41.1 | 47 | 59.6 | 27.6 | 22.1 | 35.9 |