Image Classification on Rotated MNIST (test)
98.82AccuracyGHS
Evaluation results
| Method | Links | |
|---|---|---|
| GHSModel Architecture=E(2)-equivariant model2026.03 | 98.82 | |
| APSModel Architecture=E(2)-equivariant model2026.03 | 98.12 | |
| LPFModel Architecture=E(2)-equivariant model2026.03 | 97.85 | |
| BaselineModel Architecture=E(2)-equivariant model2026.03 | 96.69 | |
| SB-CNNInput Size=28 × 282024.04 | 95.68 | |
| GD-CNNInput Size=28 × 282024.04 | 95.35 | |
| ST-CNNInput Size=28 × 282024.04 | 95.05 | |
| E(2)-CNNInput Size=29 × 292024.04 | 94.37 | |
| H-NetInput Size=32 × 322024.04 | 92.44 | |
| MAX-CNNInput Size=28 × 282024.04 | 92.33 | |
| B-CNNInput Size=32 × 322024.04 | 88.29 | |
| MAX-CNNInput Size=28 × 282024.04 | 83.32 | |
| ORNInput Size=32 × 322024.04 | 80.01 | |
| LBP-CNNInput Size=28 × 282024.04 | 75.28 | |
| RotEqNetInput Size=28 × 282024.04 | 73.2 | |
| BaselineInput Size=28 × 282024.04 | 45.35 | |
| G-CNNInput Size=28 × 282024.04 | 44.81 | |
| LBP-CNNInput Size=28 × 282024.04 | 36.31 |