List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100-C CS-3 1.0 (test)

47.33Mean AccuracyDynACL-AIR
22.494828.942435.3941.8376Apr 30, 2023Nov 9, 2023May 21, 2024Dec 1, 2024Jun 12, 2025Dec 23, 2025Jul 5, 2026
Updated 3mo ago

Evaluation results

MethodLinks
DynACL-AIRPre-training=DynACL-AIR, Finetuning procedure=AFF, Backbone=ResNet-182023.0447.33
ACL-AIRPre-training=ACL-AIR, Finetuning procedure=AFF, Backbone=ResNet-182023.0447.25
DynACLPre-training=DynACL, Finetuning procedure=AFF, Backbone=ResNet-182023.0446.71
ACLPre-training=ACL, Finetuning procedure=AFF, Backbone=ResNet-182023.0445.93
ACL-AIRPre-training=ACL-AIR, Finetuning procedure=SLF, Backbone=ResNet-182023.0441.54
ACL-AIRPre-training=ACL-AIR, Finetuning procedure=ALF, Backbone=ResNet-182023.0440.7
DynACL-AIRPre-training=DynACL-AIR, Finetuning procedure=SLF, Backbone=ResNet-182023.0439.21
ACLPre-training=ACL, Finetuning procedure=SLF, Backbone=ResNet-182023.0438.9
DynACLPre-training=DynACL, Finetuning procedure=SLF, Backbone=ResNet-182023.0438.08
FedAvg + SpecGradFilterBackbone=ResNet-20, Severity Level=3, Number of Clients=100, Client Participation Ratio (R)=0.12026.0737.04
DynACL-AIRPre-training=DynACL-AIR, Finetuning procedure=ALF, Backbone=ResNet-182023.0436.48
DynACLPre-training=DynACL, Finetuning procedure=ALF, Backbone=ResNet-182023.0435.61
ACLPre-training=ACL, Finetuning procedure=ALF, Backbone=ResNet-182023.0435.13
FedDynBackbone=ResNet-20, Severity Level=3, Number of Clients=100, Client Participation Ratio (R)=0.12026.0731.83
FedAvgBackbone=ResNet-20, Severity Level=3, Number of Clients=100, Client Participation Ratio (R)=0.12026.0726.18
SCAFFOLDBackbone=ResNet-20, Severity Level=3, Number of Clients=100, Client Participation Ratio (R)=0.12026.0725.95
FedCMBackbone=ResNet-20, Severity Level=3, Number of Clients=100, Client Participation Ratio (R)=0.12026.0723.45