Class-Incremental Learning on ImageNet-A 10 steps incremental
63.19Average AccuracySimE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SimEBackbone=CLIP ViT-B/162026.03 | 63.19 | 50.56 | |
| Boosting-CLBackbone=CLIP ViT-B/162026.03 | 61 | 49.04 | |
| APERBackbone=CLIP ViT-B/162026.03 | 60.47 | 49.37 | |
| MISABackbone=CLIP ViT-B/162026.03 | 55.32 | 47.91 | |
| DualPromptBackbone=CLIP ViT-B/162026.03 | 53.71 | 41.67 | |
| CODA-PromptBackbone=CLIP ViT-B/162026.03 | 53.54 | 42.73 | |
| L2PBackbone=CLIP ViT-B/162026.03 | 49.39 | 41.71 |