List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100 (Standard Accuracy & Latency)

84.2Top-1 AccuracySAMix
76.992878.863980.73582.6061Jun 3, 2026
Updated 5mo ago

Evaluation results

MethodLinks
SAMixBackbone=ResNet502026.0684.266.2416.68
AlignmixupBackbone=ResNet502026.0681.774.571.98
OA-CutMix/relBackbone=ResNet502026.0681.4176.281.02
PuzzleMixBackbone=ResNet502026.0681.276.23301
PuzzleMixBackbone=ResNet182026.0680.976.23301
SAMixBackbone=ResNet182026.0680.8966.2416.68
AlignmixupBackbone=ResNet182026.0680.5274.571.98
ResizeMixBackbone=ResNet182026.0680.1375.110.62
SnapMixBackbone=ResNet182026.0679.8175.7551.49
ResizeMixBackbone=ResNet502026.0679.8175.110.62
OA-CutMix/relBackbone=ResNet182026.0679.7976.281.02
MixupBackbone=ResNet502026.0679.7674.050.73
Attentive CutMixBackbone=ResNet182026.0679.3476.1656.36
CutMixBackbone=ResNet502026.0679.3475.270.61
CutMix + decoupledBackbone=ResNet502026.0679.2775.430.6
CutMixBackbone=ResNet182026.0679.2675.270.61
SnapMixBackbone=ResNet502026.0679.1875.7551.49
Attentive CutMixBackbone=ResNet502026.0679.1376.1656.36
FMixBackbone=ResNet502026.0679.1174.876.68
MixupBackbone=ResNet182026.0679.0974.050.73
CutMix + decoupledBackbone=ResNet182026.0679.0975.430.6
Mixup + CutMixBackbone=ResNet502026.0679.0873.780.67
FMixBackbone=ResNet182026.0678.9274.876.68
Mixup + CutMixBackbone=ResNet182026.0678.6473.780.67
VanillaBackbone=ResNet502026.0678.0972.14—
VanillaBackbone=ResNet182026.0677.2772.14—