Skip to main content
. 2021 Nov 12;7(4):767–782. doi: 10.3390/tomography7040064

Figure 3.

Figure 3

Architecture of the pix2pixHD model used, where the global generator consists of 3 components: a convolution front-end, a set of residual blocks, and a transposed convolutional back-end. The local generator also consists of 3 components: a convolutional front-end, a set of residual blocks, and a transposed convolutional back-end. The multiscale discriminators consist of three identically structured networks.