List your company
ResearchFormerly WizWand

Generative Modeling on CIFAR10 (test)

9.62FIDDG flow
8.8713.932518.99524.0575Dec 1, 2020
Updated 8mo ago

Evaluation results

MethodLinks
DG flowRefinement=KL, Backbone=SN-ResNet-GAN2020.129.62
DG flowRefinement=log D, Backbone=SN-ResNet-GAN2020.129.73
DG flowRefinement=JS, Backbone=SN-ResNet-GAN2020.129.79
SN-ResNet-GANRefinement=None (Base Model)2020.1214.1
DG flowRefinement=KL, Backbone=SN-DCGAN, Loss=non-saturating (ns)2020.1215.3
DG flowRefinement=KL, Backbone=SN-DCGAN, Loss=hinge (hi)2020.1215.68
DOTBackbone=SN-DCGAN, Loss=non-saturating (ns)2020.1215.78
DG flowRefinement=JS, Backbone=SN-DCGAN, Loss=non-saturating (ns)2020.1215.9
DG flowRefinement=log D, Backbone=SN-DCGAN, Loss=non-saturating (ns)2020.1216.42
DG flowRefinement=JS, Backbone=SN-DCGAN, Loss=hinge (hi)2020.1216.45
DOTBackbone=SN-DCGAN, Loss=hinge (hi)2020.1217.12
DG flowRefinement=log D, Backbone=SN-DCGAN, Loss=hinge (hi)2020.1217.36
SN-DCGANRefinement=None (Base Model), Loss=hinge (hi)2020.1220.7
SN-DCGANRefinement=None (Base Model), Loss=non-saturating (ns)2020.1220.9
DG flowRefinement=JS, Backbone=WGAN-GP2020.1223.15
DOTBackbone=WGAN-GP2020.1224.14
DG flowRefinement=log D, Backbone=WGAN-GP2020.1224.53
DG flowRefinement=KL, Backbone=WGAN-GP2020.1224.68
WGAN-GPRefinement=None (Base Model)2020.1228.37