List your company
ResearchFormerly WizWand

Image Classification on Oxford-IIIT Pet (mean per class accuracy)

0.971Mean Per Class AccuracyDom-Ad
0.8410.874750.90850.94225Sep 28, 2020Nov 2, 2020Dec 8, 2020Jan 12, 2021Feb 17, 2021Mar 24, 2021Apr 29, 2021
Updated 8mo ago

Evaluation results

MethodLinks
Dom-AdBackbone=Inception-v32020.090.971
FullBackbone=ResNet-50, Pre-training=JFT, Protocol=Full2020.090.97
Dom-AdBackbone=AmoebaNet-B2020.090.968
NNCLR + SupArch=ViT-B/82021.040.949
AdaptersBackbone=ResNet-50, Pre-training=JFT, Protocol=Adapters2020.090.946
BaselineBackbone=ResNet-502020.090.945
Sup. INArch=ViT-B/82021.040.932
NNCLRArch=ViT-B/82021.040.928
NNCLRArch=ViT-B/162021.040.925
NNCLRArch=R502021.040.918
Sup. INArch=R502021.040.915
Sup. INArch=ViT-B/162021.040.914
BYOLArch=R502021.040.904
SimCLRArch=R502021.040.846