Skip to main content
. 2023 Nov 9;13:19436. doi: 10.1038/s41598-023-46607-w

Figure 8.

Figure 8

Visual representation of the segmentation result (right) provided by the Unet++ model using the raw RGB image (left) as input in comparison to the manually labeled ground truth mask (middle). On this test image, our model achieved an Dice score of 0.94.