Incremental Learning On Cifar100 B0 10Steps
Metrics
Average Incremental Accuracy
Results
Performance results of various models on this benchmark
Model Name | Average Incremental Accuracy | Paper Title | Repository |
|---|---|---|---|
| DER(ResNet-18) | 74.64 | DER: Dynamically Expandable Representation for Class Incremental Learning | - |
| TCIL-Lite | 76.74 | Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning | - |
| Coil | 65.48 | Co-Transport for Class-Incremental Learning | - |
| TCIL | 77.30 | Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning | - |
| FOSTER | 72.9 | FOSTER: Feature Boosting and Compression for Class-Incremental Learning | - |
0 of 5 row(s) selected.