Rotated MNIST
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Image Classification | Rotated MNIST target domains 0° and 90° | Accuracy96 | 21 | |
| Continual Learning | Rotated MNIST | AutoAttack Accuracy85.64 | 10 | |
| Online Model Selection | Rotated MNIST Abrupt Drift | Cumulative Loss44.48 | 10 | |
| Image Reconstruction | Rotated MNIST SO(2) discretized (test) | Reconstruction Loss0.0162 | 6 | |
| Image Classification | Rotated MNIST (in-domain) | Accuracy98.9 | 6 | |
| Online Model Selection | Rotated MNIST Incremental Drift | Cumulative Loss42.97 | 5 | |
| Image Classification | Rotated MNIST (test) | Accuracy98.82 | 4 | |
| Image Classification | Rotated MNIST Out-of-distribution (test) | Accuracy95.8 | 3 |
Showing 8 of 8 rows