List your company
ResearchFormerly WizWand

Fine-grained Image Classification on Herbarium (test)

14.9Top-1 Test ErrorFull FT
13.90420.62727.3534.073Mar 7, 2023
Updated 8mo ago

Evaluation results

MethodLinks
Full FTBackbone=SWIN-L, % Trainable parameters=100%, No backbone backpropagation=false2023.0314.9
InCABackbone=SWIN-L, % Trainable parameters=3.7%, No backbone backpropagation=true2023.0318.2
LoRABackbone=SWIN-L, % Trainable parameters=0.8%, No backbone backpropagation=false2023.0318.4
Full FTBackbone=ViT-L/162023.0318.8
LoRABackbone=ViT-L/162023.0319.2
InCABackbone=ViT-L/162023.0321.1
AdaLNBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.0321.2
VPTBackbone=ViT-L/162023.0321.4
InCA (last)Backbone=SWIN-L, % Trainable parameters=3.7%, No backbone backpropagation=true2023.0323
InCA (last)Backbone=ViT-L/162023.0324.6
In. MLP-3Backbone=SWIN-L, % Trainable parameters=2.8%, No backbone backpropagation=true2023.0324.9
MLP-3Backbone=SWIN-L, % Trainable parameters=2.8%, No backbone backpropagation=false2023.0327.6
AdaLNBackbone=ViT-L/162023.0327.9
BitFitBackbone=ViT-L/162023.0328.3
In. LPBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=true2023.0329.2
In. MLP-3Backbone=ViT-L/162023.0329.5
BitFitBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.0329.7
In. LPBackbone=ViT-L/162023.0332.6
LPBackbone=SWIN-L, % Trainable parameters=0.1%, No backbone backpropagation=false2023.0334
MLP-3Backbone=ViT-L/162023.0336.4
LPBackbone=ViT-L/162023.0339.8