List your company
ResearchFormerly WizWand

Classification on ImageNet-16H noise level 80

6.03Error RateP+L (MAP TS)
5.6927.973510.25512.5365Sep 29, 2021
Updated 8mo ago

Evaluation results

MethodLinks
P+L (MAP TS)Backbone=VGG-19, Number of seeds=25, Labeled data points=5067-5152, Calibration=MAP TS2021.096.03
P+L (IMax Binning)Backbone=VGG-19, Number of seeds=25, Labeled data points=5067-5152, Calibration=IMax Binning2021.096.31
L+LBackbone=VGG-19, Number of seeds=25, Labeled data points=5067-51522021.096.78
P+L (Ensemble TS)Backbone=VGG-19, Number of seeds=25, Labeled data points=5067-5152, Calibration=Ensemble TS2021.096.79
P+LBackbone=GoogLeNet, Calibration=MAP TS2021.096.8
L+LBackbone=GoogLeNet2021.097.33
P+L (Uncalibrated)Backbone=VGG-19, Number of seeds=25, Labeled data points=5067-51522021.097.52
P+LBackbone=GoogLeNet, Calibration=IMax Binning2021.097.66
P+LBackbone=GoogLeNet, Calibration=Ensemble TS2021.097.73
P+LBackbone=GoogLeNet, Calibration=Uncalibrated2021.098.06
Human2021.099.99
Human2021.099.99
VGG-19Backbone=VGG-192021.0911.1
GoogLeNetBackbone=GoogLeNet2021.0914.48