Image Classification on Tiny-ImageNet (Corruption Robustness Metrics)
53.65Accuracy (Original)Zubov-Net
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Zubov-NetActivation=Tanh2025.09 | 53.65 | 40.38 | 10.14 | 39.78 | 47.33 | 45.71 | 26.33 | 29.33 | 26.84 | 33.23 | |
| SODEFActivation=sin2025.09 | 52.95 | 40.11 | 8.85 | 40.2 | 46.88 | 45.28 | 25.59 | 29.68 | 26.12 | 32.84 | |
| Neural ODEActivation=Tanh2025.09 | 52.15 | 38.26 | 9.74 | 36.29 | 42.74 | 41.44 | 22.84 | 27.03 | 20.43 | 29.85 | |
| Proj-NODEActivation=Tanh2025.09 | 51.94 | 39.03 | 8.96 | 39.63 | 46.29 | 45.15 | 26.13 | 29.16 | 24.53 | 32.36 | |
| ResNet182025.09 | 51.37 | 38.18 | 8.61 | 36.01 | 42.34 | 40.93 | 22.25 | 26.43 | 19.84 | 29.32 | |
| FxTS-Net2025.09 | 49.44 | 36.78 | 8.15 | 38.41 | 46.12 | 44.22 | 25.07 | 28.2 | 20.69 | 30.96 | |
| LyaNet2025.09 | 49.35 | 37.38 | 8.14 | 37.93 | 45.52 | 44.09 | 25.4 | 28.48 | 20.38 | 30.92 |