Visual Question Answering on Ext-VQA (test)
78.94Presence AccuracyDeepMedix-R1
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| DeepMedix-R12025.09 | 78.94 | 69.7 | 62.69 | 65.32 | 72.61 | 91.01 | 54.22 | 72.43 | |
| DeepMedix2025.09 | 75.36 | 68.39 | 62.35 | 66.58 | 75.23 | 92.59 | 46.99 | 70.79 | |
| CheXagent2025.09 | 73.46 | 62.26 | 48.08 | 35.24 | 72.21 | 78.84 | 39.16 | 62.42 | |
| Qwen2.5-VL-7Bparameters=7B, reasoning_mode=think2025.09 | 52.53 | 36.82 | 15.08 | 7.62 | 50.45 | 57.14 | 20.48 | 38.1 | |
| HuatuoGPT-Vision2025.09 | 50.93 | 49.86 | 25.18 | 20.57 | 53.47 | 49.21 | 32.53 | 43.38 | |
| Qwen2.5-VL-32Bparameters=32B2025.09 | 48.92 | 35.26 | 14.51 | 9.14 | 52.87 | 58.73 | 19.88 | 36.6 | |
| MedGemma2025.09 | 48.12 | 42.06 | 23.33 | 17.9 | 55.29 | 60.32 | 37.95 | 40.74 | |
| DeepSeek-VL-7Bparameters=7B2025.09 | 33.8 | 23.88 | 10.1 | 7.24 | 11.78 | 22.22 | 30.72 | 23.52 | |
| Qwen2.5-VL-7Bparameters=7B2025.09 | 32.4 | 33.61 | 17.64 | 10.1 | 4.83 | 51.85 | 6.02 | 26.28 | |
| LLaVA-Rad2025.09 | 6.81 | 7.25 | 2.99 | 2.29 | 1.81 | 0.53 | 0 | 5.1 |