Figure 6
From: Melanoma segmentation using deep learning with test-time augmentations and conditional random fields
Qualitative performance of proposed skin lesion segmentation technique on ISIC-2016 and ISIC-2017 datasets. The first column shows the original testing images, the actual annotated masks are shown in column 2. The third column shows the predicted mask without using any postprocessing technique. In fourth and fifth column the predicted masks refined by CRF and TTA are shown. The last column shows the predicted mask refined by using both CRF and TTA.
