Figure 2.
A schematic illustration of the CNN architectures used in current SUSAN. a) The R-Net modified from standard U-Net structure is used for the CNN mapping function F and B. The R-Net allows two outputs each of which performs image segmentation and image translation, separately. The joint portion of the R-Net for segmentation and translation branch enables sharing image features during network training. b) The discriminator CNN designed in PatchGAN is used for DX and DY. This network outputs image patches with reduced image size which will be used for differentiating real versus synthetic images in the adversarial training process. (Abbreviation: Conv: Convolution Layer; BN: Batch Normalization; ReLU: Rectified Linear Unit; LeakyReLU: Leaky Rectified Linear Unit; Deconv: Transpose Convolution Layer; Softmax: Softmax Layer)