Incremental Learning On Cifar 100 50 Classes 3

Métriques

Average Incremental Accuracy

Résultats

Résultats de performance de divers modèles sur ce benchmark

		Paper Title
TCIL	74.88	Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning
TCIL-Lite	74.30	Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning
DER(Standard ResNet-18)	72.60	DER: Dynamically Expandable Representation for Class Incremental Learning
D3Former	72.23	D3Former: Debiased Dual Distilled Transformer for Incremental Learning
FOSTER	69.46	FOSTER: Feature Boosting and Compression for Class-Incremental Learning
RMM (Modified ResNet-32)	68.86	RMM: Reinforced Memory Management for Class-Incremental Learning
DER(Modified Res-32)	67.60	DER: Dynamically Expandable Representation for Class Incremental Learning
CCIL-SD	67.17	Essentials for Class Incremental Learning
PODNet (CNN)	64.83	PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning
UCIR (CNN)*	63.42	Learning a Unified Classifier Incrementally via Rebalancing
UCIR (NME)*	63.12	Learning a Unified Classifier Incrementally via Rebalancing
iCaRL*	57.17	iCaRL: Incremental Classifier and Representation Learning
BiC	56.86	Large Scale Incremental Learning
PPCA-SWSL	-	Scalable Learning with Incremental Probabilistic PCA
PPCA-CLIP	-	Scalable Learning with Incremental Probabilistic PCA

0 of 15 row(s) selected.

Incremental Learning On Cifar 100 50 Classes 3 | SOTA | HyperAI