List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100 LT IF=10

65.5Top-1 AccuracyProCo
42.77648.675554.57560.4745Mar 11, 2024Apr 10, 2024May 10, 2024Jun 10, 2024Jul 10, 2024Aug 9, 2024Sep 9, 2024
Updated 3mo ago

Evaluation results

MethodLinks
ProCoBackbone=ResNet-32, Training Epochs=2002024.0365.5
BCLBackbone=ResNet-32, Training Epochs=2002024.0364.6
Logit AdjustmentBackbone=ResNet-32, Training Epochs=2002024.0364
Hybrid-SCBackbone=ResNet-32, Training Epochs=2002024.0363.1
ResLTBackbone=ResNet-32, Training Epochs=2002024.0362
MetaSAug-LDAMBackbone=ResNet-32, Training Epochs=2002024.0361.3
Casual ModelBackbone=ResNet-32, Training Epochs=2002024.0359.6
BBNBackbone=ResNet-32, Training Epochs=2002024.0359.1
TSCBackbone=ResNet-32, Training Epochs=2002024.0359
SSPBackbone=ResNet-32, Training Epochs=2002024.0358.9
LDAM-DRWBackbone=ResNet-32, Training Epochs=2002024.0358.7
CB-FocalBackbone=ResNet-32, Training Epochs=2002024.0358
MLLM-LLaVA-FLType=Our framework2024.0948.87
CLIP2FLType=SOTA2024.0948.2
CReFFType=Classifier-retraining2024.0947.08
CCVRType=Heterogeneity-oriented FL methods2024.0946.88
Ratio LossType=Imbalance-oriented FL methods2024.0946.79
FedNovaType=Heterogeneity-oriented FL methods2024.0946.75
FedBEType=Heterogeneity-oriented FL methods2024.0946.25
FedDFType=Heterogeneity-oriented FL methods2024.0946.19
FedProxType=Heterogeneity-oriented FL methods2024.0946.1
FedAvgType=Heterogeneity-oriented FL methods2024.0945.87
Fed-Focal LossType=Imbalance-oriented FL methods2024.0945.52
FedAvgMType=Heterogeneity-oriented FL methods2024.0944.66
FedAvg+ τ-normType=Imbalance-oriented FL methods2024.0943.65