Few-Shot Class-Incremental Learning on CIFAR-100 (60 base classes, 5-way 5-shot, Session 5)
87.35AccuracyPriViLege
Evaluation results
| Method | Links | |
|---|---|---|
| PriViLegePT. Backbone=CLIP-B/162024.05 | 87.35 | |
| CoACTPT. Backbone=ViT-B/162024.05 | 86.52 | |
| CPE-CLIPPT. Backbone=CLIP-B/162024.05 | 82.42 | |
| SV-TPT. Backbone=SwinT2024.05 | 74 | |
| BOTPT. Backbone=X2024.05 | 64.86 | |
| SoftNetPT. Backbone=X2024.05 | 61.09 | |
| SAVCPT. Backbone=X2024.05 | 59.27 | |
| FACTPT. Backbone=X2024.05 | 58.36 | |
| LIMITPT. Backbone=X2024.05 | 57.77 | |
| C-FSCILPT. Backbone=X2024.05 | 56.95 | |
| MetaFSCILPT. Backbone=X2024.05 | 56.52 | |
| CLOMPT. Backbone=X2024.05 | 56.48 | |
| CECPT. Backbone=X2024.05 | 55.57 | |
| F2MPT. Backbone=X2024.05 | 51.39 | |
| FSLL+SSPT. Backbone=X2024.05 | 49.65 | |
| FSLLPT. Backbone=X2024.05 | 48.53 | |
| IDLVQ-CPT. Backbone=X2024.05 | 47.55 | |
| TOPICPT. Backbone=X2024.05 | 35.52 | |
| iCaRLPT. Backbone=X2024.05 | 24 | |
| RebalancingPT. Backbone=X2024.05 | 21.35 |