Visual Machine Reading Comprehension on VisualMRC
307.5CIDErIXC 2.5
Evaluation Results
| Method | Links | |
|---|---|---|
| IXC 2.5Size=7B, Visual tokens=51182024.09 | 307.5 | |
| DocOwl 1.5Size=8B, Visual tokens=16982024.09 | 246.4 | |
| UReaderSize=7B, Visual tokens=8412024.09 | 221.7 | |
| DocOwl2Size=8B, Visual tokens=3242024.09 | 217.4 | |
| DocOwlSize=7B, Visual tokens=8412024.09 | 188.8 | |
| DonutSize=<1B, Visual tokens=4800, Fine-tuned=true2024.09 | 93.91 |