List your company
ResearchFormerly WizWand

Class-incremental Learning on CIFAR-100 i-Blurry-50-10 (test)

79.89Average AccuracyStandard
28.264441.667255.0768.4728Jul 31, 2025
Updated 4mo ago

Evaluation results

MethodLinks
Standard2025.0779.89
EB (w/ DMA)Number of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0759.06
EB (w/o DMA)Number of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0753.98
FOSTERNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0752.8
DPCL*Number of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0750.22
CLIBNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0749.22
EWC++Number of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0738.78
ER-MIRNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0738.28
BiCNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0737.61
GDumbNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0734.03
iCaRLNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0733.55
RMNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0733.27
FTFNumber of tasks (t)=5, Pre-trained initialization (PR)=ImageNet, i-Blurry setting (n/m)=50/102025.0730.25