Medical Image Classification on Blood
99.61Top-1 AccuracyJoint Training
Evaluation results
| Method | Links | |
|---|---|---|
| Joint Trainingreplaying prior data=false2026.03 | 99.61 | |
| Bi-CRCLreplaying prior data=false2026.03 | 98.08 | |
| CRCLreplaying prior data=false2026.03 | 97.13 | |
| MOSreplaying prior data=false2026.03 | 92.53 | |
| SLCAreplaying prior data=false2026.03 | 90.23 | |
| ADAM-Adapterreplaying prior data=false2026.03 | 88.09 | |
| FOSTERreplaying prior data=true2026.03 | 87.7 | |
| GIFBackbone=ResNet-50, Training Protocol=Trained from scratch, Diffusion Model=Fine-tuned, Number of Samples=252026.03 | 87.25 | |
| Real GuidanceBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 86.67 | |
| DERreplaying prior data=true2026.03 | 86.48 | |
| L2Preplaying prior data=false2026.03 | 86.46 | |
| SSIATreplaying prior data=false2026.03 | 86 | |
| SimpleCILreplaying prior data=false2026.03 | 83.85 | |
| iCaRLreplaying prior data=true2026.03 | 81.41 | |
| DA-FusionBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 80.76 | |
| Diff-AugBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 79.06 | |
| BaselineBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 77.67 | |
| DualPromptreplaying prior data=false2026.03 | 76.62 | |
| ACLreplaying prior data=true2026.03 | 76.09 | |
| CODAPromptreplaying prior data=false2026.03 | 72.64 | |
| Real GuidanceBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 70.67 | |
| Diff-MixBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 70.44 | |
| EASEreplaying prior data=false2026.03 | 68.85 | |
| GIFBackbone=ResNet-50, Training Protocol=Trained from scratch, Diffusion Model=Fine-tuned, Number of Samples=52026.03 | 66.58 | |
| CutMixBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 64.58 | |
| DA-FusionBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 63.91 | |
| MixupBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=252026.03 | 62.17 | |
| Diff-AugBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 61.73 | |
| CutMixBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 60.92 | |
| MixupBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 60.33 | |
| BaselineBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 60 | |
| Diff-MixBackbone=ResNet-50, Training Protocol=Trained from scratch, Number of Samples=52026.03 | 58.03 | |
| LAEreplaying prior data=false2026.03 | 55.51 | |
| Finetunereplaying prior data=false2026.03 | 37.47 |