Class Incremental Learning On Cifar100
Métriques
10-stage average accuracy
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | 10-stage average accuracy |
|---|---|
| icarl-incremental-classifier-and | 63.24 |
| efficient-lifelong-learning-with-a-gem | 45.76 |
| continual-learning-with-tiny-episodic | 48.66 |
| learning-a-unified-classifier-incrementally | 56.53 |
| always-be-dreaming-a-new-approach-for-data | 54.44 |
| split-and-bridge-adaptable-class-incremental | 68.18 |
| supervised-contrastive-learning | 65.98 |