[8eeb5a]: / training_method_table

Download this file

37 lines (35 with data), 1.2 kB

\toprule
Model & No Augmentation & Vanilla Augmentation & Consistency Training\\
\toprule
\multicolumn{4}{c}{\textbf{Kvasir-SEG }}\\
\midrule

        DD-DeepLabV3+& 0.829 & 0.848 & 0.852 \\
        DeepLab& 0.822 & 0.850 & 0.852 \\
        FPN& 0.822 & 0.853 & 0.852 \\
        TriUnet& 0.817 & 0.841 & 0.845 \\
        Unet& 0.828 & 0.851 & 0.851 \\
\midrule
\multicolumn{4}{c}{\textbf{Etis-LaribDB}}
\midrule
        DD-DeepLabV3+& 0.408 & 0.460 & 0.482 \\
        DeepLab& 0.417 & 0.472 & 0.505 \\
        FPN& 0.404 & 0.440 & 0.475 \\
        TriUnet& 0.309 & 0.410 & 0.434 \\
        Unet& 0.403 & 0.447 & 0.481 \\
\midrule
\multicolumn{4}{c}{\textbf{CVC-ClinicDB}}
\midrule
        DD-DeepLabV3+ & 0.681 & 0.728 0.736 \\
        DeepLab & 0.684 & 0.733 0.740 \\
        FPN& 0.675 & 0.715  0.727 \\
        TriUnet& 0.623 & 0.684 0.696 \\
        Unet& 0.679 & 0.717 & 0.730 \\
\midrule
\multicolumn{4}{c}{\textbf{EndoCV2020}}\\
\midrule

        DD-DeepLabV3+& 0.596 & 0.668 & 0.668 \\
        DeepLab& 0.608 & 0.676 & 0.676 \\
        FPN& 0.600 & 0.662 & 0.673 \\
        TriUnet& 0.577 & 0.667 & 0.684 \\
        Unet& 0.598 & 0.660 & 0.676 \\
\bottomrule