Skin Lesion Classification on HAM10000
96.32Accuracysoft voting ensemble of CNNs
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| soft voting ensemble of CNNsBackbone=MobileNetV22025.12 | 96.32 | — | — | |
| Ji et al.2025.12 | 94.31 | — | — | |
| Adebiyi et al.2025.12 | 94.11 | — | — | |
| DermViT (ViT)Year=2024, Classes=4, XAI=Limited2026.01 | 92.48 | 91.2 | 0.98 | |
| Ours (EfficientNetV2-L + Attention)Year=2025, Classes=7, XAI=Yes2026.01 | 91.15 | 91.13 | 0.993 | |
| Zhang et al. (Vision Transformer)Year=2023, Classes=7, XAI=Partial2026.01 | 90.2 | 88.7 | 0.981 | |
| Thwin et al.2025.12 | 90 | — | — | |
| Gwilliam et al. (ResNet-50 + XAI)Year=2023, Classes=7, XAI=Yes2026.01 | 89.5 | 87.9 | 0.978 | |
| Tschandl et al. (Ensemble CNN)Year=2020, Classes=7, XAI=No2026.01 | 88.9 | 87.2 | 0.976 |