List your company
ResearchFormerly WizWand

Image Classification on ImageNet 64x64 (non-iid)

38.7AccuracyStandard
-1.4448.97819.429.822May 26, 2023
Updated 8mo ago

Evaluation results

MethodLinks
StandardBackbone=ResNet56, SFD/FedRolex ratio=1.0, Training rounds=40002023.0538.7
SLTBackbone=ResNet56, SFD/FedRolex ratio=0.5, Training rounds=40002023.0531.8
SLTBackbone=ResNet56, SFD/FedRolex ratio=0.25, Training rounds=40002023.0529.7
Small modelBackbone=ResNet56, SFD/FedRolex ratio=0.5, Training rounds=40002023.0527.3
SLTBackbone=ResNet56, SFD/FedRolex ratio=0.125, Training rounds=40002023.0521.7
FedRolexBackbone=ResNet56, SFD/FedRolex ratio=0.5, Training rounds=40002023.0518.4
Small modelBackbone=ResNet56, SFD/FedRolex ratio=0.25, Training rounds=40002023.0516.2
FDBackbone=ResNet56, SFD/FedRolex ratio=0.5, Training rounds=40002023.0515.7
FedRolexBackbone=ResNet56, SFD/FedRolex ratio=0.25, Training rounds=40002023.0510.2
Small modelBackbone=ResNet56, SFD/FedRolex ratio=0.125, Training rounds=40002023.058.4
FedRolexBackbone=ResNet56, SFD/FedRolex ratio=0.125, Training rounds=40002023.052.8
FDBackbone=ResNet56, SFD/FedRolex ratio=0.125, Training rounds=40002023.050.1
FDBackbone=ResNet56, SFD/FedRolex ratio=0.25, Training rounds=40002023.050.1