Medical Visual Question Answering on OVQA
79.58AccuracyKG-CMI
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| KG-CMIreported_format=Mean ± Standard Deviation2026.04 | 79.58 | 71.77 | 84.47 | |
| M3AEreported_format=Mean ± Standard Deviation2026.04 | 75.32 | 61.65 | 84.37 | |
| VG-CALFreported_format=Mean ± Standard Deviation2026.04 | 74.21 | 60.74 | 83.96 | |
| LaPAreported_format=Mean ± Standard Deviation2026.04 | 72.51 | 56.46 | 83.12 | |
| MQATreported_format=Mean ± Standard Deviation2026.04 | 67 | 38.21 | 79.92 | |
| MMQ*reported_format=Point estimate2026.04 | 64.14 | 46.04 | 76.13 | |
| VQAMix*reported_format=Point estimate2026.04 | 64.14 | 44.98 | 76.83 | |
| MUMCreported_format=Mean ± Standard Deviation2026.04 | 61.2 | 39.05 | 75.87 | |
| MEVF*reported_format=Point estimate2026.04 | 58.46 | 34.69 | 74.21 | |
| MedCausalXEvaluation mode=Zero-shot2026.03 | 58.4 | — | — | |
| M2I2reported_format=Mean ± Standard Deviation2026.04 | 57.89 | 36.02 | 72.38 | |
| MPRreported_format=Mean ± Standard Deviation2026.04 | 56.97 | 31.7 | 72.6 | |
| MedVLM-R1Evaluation mode=Zero-shot2026.03 | 50.3 | — | — | |
| InternVL-2Evaluation mode=Zero-shot2026.03 | 49.8 | — | — | |
| LLaVA-MedEvaluation mode=Zero-shot2026.03 | 47.2 | — | — | |
| Med-R1Evaluation mode=Zero-shot2026.03 | 44.8 | — | — | |
| GPT-4oEvaluation mode=Zero-shot2026.03 | 43.5 | — | — |