List your company
ResearchFormerly WizWand

Image Classification on iNaturalist 2017

83.8AccuracyHiera-H448
58.00864.70471.478.096Jul 17, 2021Nov 8, 2021Mar 2, 2022Jun 24, 2022Oct 16, 2022Feb 7, 2023Jun 1, 2023
Updated 8mo ago

Evaluation results

MethodLinks
Hiera-H448backbone=Hiera-H4482023.0683.8
MAEBackbone=ViT-H, Input Resolution=448, Pre-train Data=IN1K, Evaluation Protocol=Fine-tuned2021.1183.4
ViT-H448backbone=ViT-H4482023.0683.4
Hiera-Hbackbone=Hiera-H2023.0679.6
MAEBackbone=ViT-H, Pre-train Data=IN1K, Evaluation Protocol=Fine-tuned2021.1179.3
ViT-Hbackbone=ViT-H2023.0679.3
Hiera-Lbackbone=Hiera-L2023.0676.8
MAEBackbone=ViT-L, Pre-train Data=IN1K, Evaluation Protocol=Fine-tuned2021.1175.7
ViT-Lbackbone=ViT-L2023.0675.7
Previous best2021.1175.4
Hiera-B+backbone=Hiera-B+2023.0674.7
Hiera-Bbackbone=Hiera-B2023.0673.3
MAEBackbone=ViT-B, Pre-train Data=IN1K, Evaluation Protocol=Fine-tuned2021.1170.5
ViT-Bbackbone=ViT-B2023.0670.5
RAMS-TransBackbone=ViT-B_162021.0768.5
TASNBackbone=ResNet1012021.0768.2
IncResNetV2Backbone=InResNetV22021.0767.3
ViTBackbone=ViT-B_162021.0767
Huang et al.Backbone=ResNet1012021.0766.8
TransFG & PSMBackbone=ViT-B_162021.0766.6
SSNBackbone=ResNet1012021.0765.2
Resnet152Backbone=ResNet1522021.0759