List your company
ResearchFormerly WizWand

Image Classification on CIFAR-100 40% symmetric noise (test)

75.91AccuracyFaMUS
45.011653.033361.05569.0767Apr 30, 2021May 6, 2021May 13, 2021May 19, 2021May 26, 2021Jun 1, 2021Jun 8, 2021
Updated 8mo ago

Evaluation results

MethodLinks
FaMUSBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0475.91——
DivideMixBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0475.2——
MentorMixBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0471.3——
M-correctionBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0470.1——
MW-NetBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0468.11——
MentorNetBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0466.8——
MixupBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0466.8——
L2RBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0462.81——
Negative Label SmoothingSmooth Rate=-42021.0658.35——
Negative Label SmoothingSmooth Rate=-22021.0658.1——
Negative Label SmoothingSmooth Rate=-62021.0657.73——
Negative Label SmoothingSmooth Rate=-12021.0657.44——
Negative Label SmoothingSmooth Rate=-0.82021.0657.35——
Negative Label SmoothingSmooth Rate=-0.42021.0656.87——
Label SmoothingSmooth Rate=0.42021.0654.26——
Label SmoothingSmooth Rate=0.82021.0654.04——
Label SmoothingSmooth Rate=0.62021.0652.72——
Vanilla LossSmooth Rate=02021.0648.03——
Co-teachingBackbone=PreAct ResNet-18, Noise Type=Symmetric, Noise Rate=40%2021.0446.2——
CEFine-tuning Protocol=FFT, Backbone=CLIP2024.09—84.0682.6
DEFTFine-tuning Protocol=Ours, Backbone=CLIP2024.09—88.1788.11
ELRFine-tuning Protocol=FFT, Backbone=CLIP2024.09—83.6683.66
GMMFine-tuning Protocol=FFT, Backbone=CLIP2024.09—87.2187.21
SCEFine-tuning Protocol=FFT, Backbone=CLIP2024.09—83.8483.84