Medical Image Classification on INbreast
82.44AccuracyQwen2.5-VL-3B+CoT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5-VL-3B+CoTTraining Data=MammoExpert, Chain-of-Thought=true2026.06 | 82.44 | — | 83.1 | |
| DKGH-MoEBackbone=ResNet-18, Sparsity=Sparse2026.01 | 79.12 | 80.84 | — | |
| Qwen2.5-VL-3BTraining Data=MammoExpert2026.06 | 78.26 | — | 78.95 | |
| DKGH-MoEBackbone=ResNet-18, Sparsity=Dense2026.01 | 77.79 | 82.91 | — | |
| DKGH-MoEBackbone=ResNet-50, Sparsity=Dense2026.01 | 77.23 | 84.01 | — | |
| EfficientMammoTraining Data=Standard2026.06 | 77.12 | 82.26 | 77.84 | |
| Qwen2.5-VL-3B+CoTTraining Data=Standard, Chain-of-Thought=true2026.06 | 76.83 | — | 77.54 | |
| Med-FlamingoTraining Data=Standard2026.06 | 76.54 | 85.65 | 77.26 | |
| MedVInTTraining Data=Standard2026.06 | 75.84 | 84.8 | 76.59 | |
| Qwen2.5-VL-3BTraining Data=Standard2026.06 | 75.52 | — | 76.32 | |
| DKGH-MoEBackbone=ResNet-50, Sparsity=Sparse2026.01 | 75.46 | 81.07 | — | |
| PLIPTraining Data=Standard2026.06 | 75.12 | 84.15 | 75.88 | |
| DenseMammo-MultiTraining Data=Standard, View=Multi2026.06 | 75.06 | 82.56 | 75.79 | |
| BioViLTraining Data=Standard2026.06 | 73.96 | 83.4 | 74.83 | |
| DenseMammo-SingleTraining Data=Standard, View=Single2026.06 | 73.14 | 80.52 | 74.01 | |
| DE-MoEBackbone=ResNet-18, Sparsity=Sparse2026.01 | 72.77 | 68.88 | — | |
| MedCLIPTraining Data=Standard2026.06 | 72.32 | 82.85 | 73.14 | |
| DD-MoEBackbone=ResNet-18, Sparsity=Dense2026.01 | 70.45 | 72.46 | — | |
| DE-MoEBackbone=ResNet-50, Sparsity=Sparse2026.01 | 66.38 | 62.5 | — | |
| DE-MoEBackbone=ResNet-18, Sparsity=Dense2026.01 | 65.02 | 72.03 | — | |
| baselineBackbone=ResNet-18, Sparsity=Dense2026.01 | 62.92 | 70.45 | — | |
| DD-MoEBackbone=ResNet-18, Sparsity=Sparse2026.01 | 62.62 | 68.1 | — | |
| DE-MoEBackbone=ResNet-50, Sparsity=Dense2026.01 | 58.2 | 58.46 | — | |
| DD-MoEBackbone=ResNet-50, Sparsity=Sparse2026.01 | 57.04 | 59.02 | — | |
| baselineBackbone=ResNet-50, Sparsity=Dense2026.01 | 56 | 67.42 | — | |
| DD-MoEBackbone=ResNet-50, Sparsity=Dense2026.01 | 54.85 | 61.46 | — | |
| MorphHRTraining Data=Standard2026.06 | — | 81.38 | — | |
| MorphResTraining Data=Standard2026.06 | — | 81.22 | — | |
| VGGMammoTraining Data=Standard2026.06 | — | 73.52 | — |