Medical Image Classification on Kvasir
92.84AccuracyFedFew
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| FedFewSetting=Personalized Methods (Multiple Server Models)2026.03 | 92.84 | — | — | — | 83.9 | 99.77 | |
| FedAMPSetting=Personalized Methods (Single Model per Client)2026.03 | 92.76 | — | — | — | 82.2 | 100 | |
| FedRepSetting=Personalized Methods (Single Model per Client)2026.03 | 92.71 | — | — | — | 82.93 | 100 | |
| FedMTLSetting=Centralized Methods (Single Global Model)2026.03 | 92.46 | — | — | — | 82.2 | 100 | |
| DittoSetting=Personalized Methods (Single Model per Client)2026.03 | 92.37 | — | — | — | 80.73 | 100 | |
| Local-onlySetting=Local-only Baseline2026.03 | 92.16 | — | — | — | 80.49 | 100 | |
| APFLSetting=Personalized Methods (Single Model per Client)2026.03 | 91.97 | — | — | — | 82.2 | 99.77 | |
| FedAvgSetting=Centralized Methods (Single Global Model)2026.03 | 85.96 | — | — | — | 82.2 | 89.2 | |
| MKG-CARECategory=Ours2026.05 | 84 | 98.4 | — | — | — | — | |
| KEMCategory=Knowledge-enhanced, Year=20252026.05 | 82.3 | 97.5 | — | — | — | — | |
| IFCASetting=Personalized Methods (Multiple Server Models)2026.03 | 82.05 | — | — | — | 40.24 | 100 | |
| Gemini-3-proCategory=General LVLM, Year=20252026.05 | 81.6 | 87.1 | — | — | — | — | |
| SwinV2-B + PHG + multi-scale + multi-filtrationsBackbone=SwinV2-B, PHG=true, multi-scale=true, multi-filtrations=true2025.12 | 81.52 | 97.25 | 81.52 | 97.4 | — | — | |
| SENet154 + PHG + multi-scale + multi-filtrationsBackbone=SENet154, PHG=true, multi-scale=true, multi-filtrations=true2025.12 | 80.42 | 98.11 | 80.42 | 97.43 | — | — | |
| SwinV2-B + PHG + multi-scaleBackbone=SwinV2-B, PHG=true, multi-scale=true, multi-filtrations=false2025.12 | 80.36 | 97.15 | 80.36 | 97.27 | — | — | |
| GPT-5.1Category=General LVLM, Year=20252026.05 | 80.3 | 85.2 | — | — | — | — | |
| FedProxSetting=Centralized Methods (Single Global Model)2026.03 | 79.91 | — | — | — | 57.51 | 86.58 | |
| SwinV2-B + PHGBackbone=SwinV2-B, PHG=true, multi-scale=false, multi-filtrations=false2025.12 | 79.78 | 97.05 | 79.78 | 97.09 | — | — | |
| ResNet152 + PHG + multi-scale + multi-filtrationsBackbone=ResNet152, PHG=true, multi-scale=true, multi-filtrations=true2025.12 | 79.52 | 97.63 | 79.52 | 97.88 | — | — | |
| SENet154 + PHG + multi-scaleBackbone=SENet154, PHG=true, multi-scale=true, multi-filtrations=false2025.12 | 79.33 | 98.01 | 79.33 | 97.19 | — | — | |
| MedMambaCategory=Vision-based, Year=20242026.05 | 79.3 | 97.6 | — | — | — | — | |
| Adaptive Saliency-Guided Robust Quantizationsaliency masking ratio=50%, batch size=128, epochs=502025.12 | 79.05 | — | 84.37 | 84.37 | — | — | |
| SENet154 + PHGBackbone=SENet154, PHG=true, multi-scale=false, multi-filtrations=false2025.12 | 79 | 97.85 | 79 | 97 | — | — | |
| SwinV2-BBackbone=SwinV2-B, PHG=false, multi-scale=false, multi-filtrations=false2025.12 | 78.58 | 97.35 | 78.58 | 97.08 | — | — | |
| ResNet152 + PHG + multi-scaleBackbone=ResNet152, PHG=true, multi-scale=true, multi-filtrations=false2025.12 | 78.27 | 97.18 | 78.27 | 97.7 | — | — | |
| SENet154Backbone=SENet154, PHG=false, multi-scale=false, multi-filtrations=false2025.12 | 78 | 97.45 | 78 | 96.5 | — | — | |
| ResNet152 + PHGBackbone=ResNet152, PHG=true, multi-scale=false, multi-filtrations=false2025.12 | 77.52 | 97.03 | 77.52 | 97.59 | — | — | |
| Saliency-Guided Training (SGT Model)saliency masking ratio=50%, batch size=128, epochs=502025.12 | 77.48 | — | 78.12 | 78.12 | — | — | |
| EfficientNetV2Category=Vision-based, Year=20212026.05 | 76.3 | 96.3 | — | — | — | — | |
| ResNet152Backbone=ResNet152, PHG=false, multi-scale=false, multi-filtrations=false2025.12 | 75.92 | 96.33 | 75.92 | 96.99 | — | — | |
| Swin-TransformerCategory=Vision-based, Year=20212026.05 | 75.2 | 97 | — | — | — | — | |
| ConvNeXtCategory=Vision-based, Year=20222026.05 | 72.5 | 96.6 | — | — | — | — | |
| VMambaCategory=Vision-based, Year=20242026.05 | 69.8 | 96.2 | — | — | — | — | |
| Qwen3.5-plusCategory=General LVLM, Year=20262026.05 | 63.4 | 79.1 | — | — | — | — | |
| HuatuoGPT-Vision-34BCategory=Med-LVLM, Year=20242026.05 | 47.3 | 69.9 | — | — | — | — | |
| Health-GPT-XL32Category=Med-LVLM, Year=20242026.05 | 34.5 | 62.6 | — | — | — | — | |
| medgemma-1.5-4b-itCategory=Med-LVLM, Year=20262026.05 | 29.2 | 59.8 | — | — | — | — |