List your company
ResearchFormerly WizWand

Image Classification on CIFAR-10-LT λ=100 (test)

19.46Error RateMetaSAug with cross-entropy loss
18.996422.125725.25528.3843Dec 15, 2021Sep 3, 2022May 24, 2023Feb 10, 2024Oct 30, 2024Jul 19, 2025Apr 8, 2026
Updated 6mo ago

Evaluation results

MethodLinks
MetaSAug with cross-entropy lossBackbone=ResNet-322021.1219.46
Bag of TricksBackbone=ResNet-322021.1219.97
RISDABackbone=ResNet-322021.1220.11
BBNBackbone=ResNet-322021.1220.18
Hybrid-PSCBackbone=ResNet-322021.1221.18
Meta-class-weight with cross-entropy lossBackbone=ResNet-322021.1223.59
GCEModel=Wide-ResNet2026.0423.75
GCEModel=DenseNet1212026.0423.79
MMCEModel=Wide-ResNet2026.0425.01
CEModel=Wide-ResNet2026.0425.2
Brier LossModel=DenseNet1212026.0425.3
GCEModel=ResNet-502026.0425.35
GCEModel=ResNet-1102026.0425.62
DFLModel=Wide-ResNet2026.0425.63
CEModel=DenseNet1212026.0426.02
MMCEModel=DenseNet1212026.0426.07
DFLModel=DenseNet1212026.0426.21
Meta-weight netBackbone=ResNet-322021.1226.43
CEModel=ResNet-502026.0426.62
MMCEModel=ResNet-502026.0426.69
MMCEModel=ResNet-1102026.0426.77
FLSDModel=DenseNet1212026.0426.84
CEModel=ResNet-1102026.0426.85
FLSDModel=Wide-ResNet2026.0426.88
DFLModel=ResNet-502026.0426.89
MixupBackbone=ResNet-322021.1226.94
DFLModel=ResNet-1102026.0427
Class-balanced CEBackbone=ResNet-322021.1227.32
Brier LossModel=Wide-ResNet2026.0427.58
L2RWBackbone=ResNet-322021.1227.77
FLSDModel=ResNet-1102026.0428.16
FLSDModel=ResNet-502026.0428.43
Class-balanced fine-tuningBackbone=ResNet-322021.1228.66
AFLModel=Wide-ResNet2026.0428.79
AFLModel=DenseNet1212026.0428.92
AFLModel=ResNet-1102026.0429.24
Cross-EntropyBackbone=ResNet-322021.1229.86
AFLModel=ResNet-502026.0429.92
Brier LossModel=ResNet-1102026.0430.82
Brier LossModel=ResNet-502026.0431.05