Fig. 9
From: Self-supervised model-informed deep learning for low-SNR SS-OCT domain transformation
Comparison results among different methods and data sets. For datasets SS-OCT, 3D-OCT1000 and Bioptigen, we used 60 reference images and for A2A using 18 images for ROI selection. First 2 rows show value of metrics over each dataset. KID metric to show how much output of each method can be realistzic.
