Image Classification on Spherical MNIST NR/NR
99.45AccuracyHexRUnet
Evaluation results
| Method | Links | |
|---|---|---|
| HexRUnetEquivariant=false2023.07 | 99.45 | |
| Gauge CNNEquivariant=true2023.07 | 99.43 | |
| SWSCNNparams=58k2020.06 | 99.37 | |
| UGSCNNEquivariant=false2023.07 | 99.23 | |
| HEAL-SWINEquivariant=false2023.07 | 99.2 | |
| Planar CNNparams=59k2020.06 | 99.07 | |
| Esteves et al.params=57k2020.06 | 98.75 | |
| SphCNNEquivariant=true2023.07 | 98.75 | |
| Kondor et al.2020.06 | 96.4 | |
| Clebsch-Gordan NetsEquivariant=true2023.07 | 96.4 | |
| S2CNNEquivariant=true2023.07 | 96 | |
| Cohen et al.params=58k2020.06 | 95.59 |