Image Classification on STL10-scale non-ideal downsampling
72.62AccuracyScale-Equivariant Fourier Networks
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Scale-Equivariant Fourier NetworksResizing strategy=zero-padding2023.11 | 72.62 | 56.24 | 0.06 | |
| Fourier CNNResizing strategy=zero-padding2023.11 | 53.57 | 53.12 | 0.2 | |
| SI-CovNetResizing strategy=zero-padding2023.11 | 52.34 | 39.58 | 0.07 | |
| SESNResizing strategy=zero-padding2023.11 | 51.55 | 46.87 | 0.07 | |
| DISCOResizing strategy=zero-padding2023.11 | 51.25 | 44.79 | 0.12 | |
| DSSResizing strategy=zero-padding2023.11 | 47.56 | 36.45 | 0.03 | |
| Wide ResNetResizing strategy=zero-padding2023.11 | 44.56 | 32.29 | 1.08 | |
| SS-CNNResizing strategy=zero-padding2023.11 | 34.18 | 21.87 | 1.72 |