Exemplar-Free Class-Incremental Learning on ImageNet-R 5-task (test)
49.34Average Accuracy (AA)Inf-SSM
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Inf-SSMBackbone=Vim-Small, Regularization focus=geometry-aware2025.05 | 49.34 | 67.51 | 25.14 | |
| SIBackbone=Vim-Small, Regularization focus=(A, B, C)2025.05 | 45.72 | 65.18 | 47.21 | |
| EWCBackbone=Vim-Small, Regularization focus=(A, B, C)2025.05 | 45.58 | 65.62 | 47.31 | |
| LwF-ABCBackbone=Vim-Small, Regularization focus=(A, B, C)2025.05 | 45.09 | 65.69 | 40.77 | |
| MASBackbone=Vim-Small, Regularization focus=(A, B, C)2025.05 | 44.7 | 65.59 | 48.23 | |
| Inf-SSMBackbone=Vim-tiny2025.05 | 39.85 | 58.74 | 23.33 | |
| SeqBackbone=Vim-Small, Regularization focus=(A, B, C)2025.05 | 38.36 | 61.29 | 56.43 | |
| LwF-ABCBackbone=Vim-tiny, Regularization focus=parameter sets (A, B, C)2025.05 | 37.17 | 58.87 | 35.1 | |
| EWCBackbone=Vim-tiny, Regularization focus=parameter sets (A, B, C)2025.05 | 35.73 | 57.62 | 48.34 | |
| MASBackbone=Vim-tiny, Regularization focus=parameter sets (A, B, C)2025.05 | 27.94 | 52.47 | 60.15 | |
| SeqBackbone=Vim-tiny, Regularization focus=parameter sets (A, B, C)2025.05 | 25.68 | 50.28 | 63.05 | |
| SIBackbone=Vim-tiny, Regularization focus=parameter sets (A, B, C)2025.05 | 25.55 | 50.05 | 63.38 |