Image Classification on Imagenette (top-1 and top-5)
85.96Top-1 AccuracyMEDiC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MEDiCBackbone=ViT-B, Epochs=300, Evaluation Protocol=kNN (frozen)2026.03 | 85.96 | 98.47 | |
| CMAEBackbone=ViT-B, Epochs=800, Evaluation Protocol=kNN (frozen)2026.03 | 77.91 | 96.33 | |
| MaskDistillBackbone=ViT-B, Epochs=300, Evaluation Protocol=kNN (frozen)2026.03 | 76 | 95.77 | |
| BootMAEBackbone=ViT-B, Epochs=800, Evaluation Protocol=kNN (frozen)2026.03 | 70.7 | 94.22 | |
| SemMAEBackbone=ViT-B, Epochs=800, Evaluation Protocol=kNN (frozen)2026.03 | 70.04 | 93.55 | |
| MAEBackbone=ViT-B, Epochs=800, Evaluation Protocol=kNN (frozen)2026.03 | 54.75 | 89.35 | |
| SimMIMBackbone=ViT-B, Epochs=400, Evaluation Protocol=kNN (frozen)2026.03 | 51.44 | 87.21 | |
| BEiTBackbone=ViT-B, Epochs=800, Evaluation Protocol=kNN (frozen)2026.03 | 30.06 | 75.69 |