List your company

Semantic Segmentation on ISIC 2018 (test)

92.05DSCCLIPSeg DA
63.60670.990578.37585.7595Oct 18, 2018Sep 21, 2019Aug 25, 2020Jul 29, 2021Jul 3, 2022Jun 6, 2023May 10, 2024

Evaluation results

MethodLinks
CLIPSeg DATraining Strategy=Adapter fine-tuning, Trainable Parameters=3M2024.0592.05—————85.9854.38
CLIPSegTraining Strategy=E2E fine-tuning, Trainable Parameters=150M2024.0591.88—————85.7660.93
CRISTraining Strategy=E2E fine-tuning, Trainable Parameters=147M2024.0591.49—————85.4164.39
CLIPSeg SATraining Strategy=Adapter fine-tuning, Trainable Parameters=4.2M2024.0591.4—————85.0560.29
SANTraining Strategy=Adapter fine-tuning, Trainable Parameters=8.4M2024.0590.39—————83.6187.25
Attn U-Net + Multi-Input + FTLParameters=α = 0.7, β = 0.3, γ = 4/3, Loss Function=Focal Tversky Loss, Architecture=Input Pyramid2018.1085.685.889.7—————
BCDU-Netd=32019.0885.1—78.598.293.70.92893.7—
BCDU-Netd=12019.0884.7—78.39893.60.92293.6—
Attn U-Net + Multi-Input + TLParameters=α = 0.7, β = 0.3, Loss Function=Tversky Loss, Architecture=Input Pyramid2018.1084.182.391.2—————
U-Net + TLParameters=α = 0.7, β = 0.3, Loss Function=Tversky Loss2018.1083.882.291.7—————
U-Net + FTLParameters=α = 0.7, β = 0.3, γ = 4/3, Loss Function=Focal Tversky Loss2018.1082.979.792.6—————
Attn U-Net + Multi-Input + DLParameters=α = 0.5, β = 0.5, Loss Function=Dice Loss, Architecture=Input Pyramid2018.1082.789.682.9—————
U-Net + DLParameters=α = 0.5, β = 0.5, Loss Function=Dice Loss2018.108284.986.7—————
Attn U-Net + DLParameters=α = 0.5, β = 0.5, Loss Function=Dice Loss2018.1080.687.482.7—————
Attention R2U-Net2019.0869.1—72.697.190.40.82259.2—
R2U-net2019.0867.9—79.292.8880.74158.1—
Attention U-net2019.0866.5—71.796.789.70.78756.6—
U-net2019.0864.7—70.896.4890.77954.9—