Open-ended image description on GPT-4o assisted evaluation
8.76AccuracyR-CoV
Evaluation results
| Method | Links | ||
|---|---|---|---|
| R-CoVModel=Qwen2.5-VL2026.04 | 8.76 | 9.72 | |
| VanillaModel=Qwen2.5-VL2026.04 | 8.35 | 9.65 | |
| R-CoVModel=LLaVA-1.52026.04 | 7.48 | 9.03 | |
| R-CoVModel=MiniGPT-42026.04 | 7.24 | 8.9 | |
| R-CoVModel=mPLUG-Owl2026.04 | 6.91 | 8.41 | |
| VanillaModel=LLaVA-1.52026.04 | 6.48 | 8.91 | |
| VanillaModel=MiniGPT-42026.04 | 5.69 | 8.62 | |
| VanillaModel=mPLUG-Owl2026.04 | 4.78 | 7.94 |