Class Incremental Learning on S-Tiny-ImageNet
59.99AccuracyJOINT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| JOINTBackbone=ResNet-18, Pre-training=None, Batch size=322024.10 | 59.99 | — | |
| CLFD-CLS-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 18.73 | — | |
| CLS-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 16.35 | — | |
| CLFD-ER-ACEBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 15.88 | — | |
| CLFD-DER++Backbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 15.42 | — | |
| CLFD-CLS-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 12.61 | — | |
| DER++Backbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 11.29 | — | |
| CLS-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 11.09 | — | |
| ER-ACEBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 11.09 | — | |
| CLFD-ER-ACEBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 10.71 | — | |
| CLFD-DER++Backbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 10.69 | — | |
| CLFD-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 8.89 | — | |
| LwFBackbone=ResNet-18, Pre-training=None, Batch size=322024.10 | 8.46 | — | |
| ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=1252024.10 | 8.3 | — | |
| ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 8.14 | — | |
| DER++Backbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 8 | — | |
| SGDBackbone=ResNet-18, Pre-training=None, Batch size=322024.10 | 7.92 | — | |
| CLFD-ERBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 7.61 | — | |
| oEWCBackbone=ResNet-18, Pre-training=None, Batch size=322024.10 | 7.58 | — | |
| ER-ACEBackbone=ResNet-18, Pre-training=None, Batch size=32, Buffer size=502024.10 | 6.68 | — | |
| SIBackbone=ResNet-18, Pre-training=None, Batch size=322024.10 | 6.58 | — |