Grounded Diagnostic Segmentation on DiagSeg
58.9DiagnosisMEDSIGHT
Evaluation Results
| Method | Links | |
|---|---|---|
| MEDSIGHTType=Unified, # Params=8B, # Data=72K, Evaluation Protocol=Zero-shot inference2026.06 | 58.9 | |
| Lingshu-7BType=Comprehension only, # Params=7B, # Data=7.1M, Evaluation Protocol=Partially trained on evaluation benchmarks2026.06 | 55.4 | |
| HealthGPT-L14Type=Comprehension only, # Params=14B, # Data=1.5M, Evaluation Protocol=Partially trained on evaluation benchmarks2026.06 | 51.2 | |
| HuatuoGPT-VisionType=Comprehension only, # Params=7B, # Data=647K, Evaluation Protocol=Zero-shot inference2026.06 | 51.1 | |
| Llama-3.2Type=Comprehension only, # Params=11B, # Data=-, Evaluation Protocol=Zero-shot inference2026.06 | 46.4 | |
| LLaVA-v1.5Type=Comprehension only, # Params=7B, # Data=158K, Evaluation Protocol=Zero-shot inference2026.06 | 45.3 | |
| InternVL3.5Type=Comprehension only, # Params=8B, # Data=16.3M, Evaluation Protocol=Partially trained on evaluation benchmarks2026.06 | 44.6 | |
| Qwen3-VLType=Comprehension only, # Params=8B, # Data=-, Evaluation Protocol=Partially trained on evaluation benchmarks2026.06 | 44.5 | |
| HealthGPT-M3Type=Comprehension only, # Params=3.8B, # Data=1.5M, Evaluation Protocol=Partially trained on evaluation benchmarks2026.06 | 44.1 | |
| InternVL2Type=Comprehension only, # Params=8B, # Data=7.3M, Evaluation Protocol=Zero-shot inference2026.06 | 42.3 | |
| InstructBLIPType=Comprehension only, # Params=7B, # Data=364K, Evaluation Protocol=Zero-shot inference2026.06 | 35.9 | |
| LLaVA-MedType=Comprehension only, # Params=7B, # Data=60K, Evaluation Protocol=Zero-shot inference2026.06 | 35.3 | |
| Med-FlamingoType=Comprehension only, # Params=8.3B, # Data=1.3M, Evaluation Protocol=Zero-shot inference2026.06 | 34.6 | |
| Yi-VLType=Comprehension only, # Params=6B, # Data=10K, Evaluation Protocol=Zero-shot inference2026.06 | 33 | |
| OMG-LLaVAType=Unified, # Params=7B, # Data=1.2M, Evaluation Protocol=Zero-shot inference2026.06 | 28 | |
| BLIP-2Type=Comprehension only, # Params=6.7B, # Data=-, Evaluation Protocol=Zero-shot inference2026.06 | 22.3 | |
| MedPLIBType=Unified, # Params=14B/7B, # Data=500K, Evaluation Protocol=Zero-shot inference2026.06 | 13.1 |