In contrast to typical GANs, a U-Net GAN uses a segmentation network as the discriminator. This segmentation network predicts two classes: real and fake. In doing so, the discriminator gives the generator region-specific feedback. This discriminator design also enables a CutMix-based consistency regularization on the two-dimensional output of the U-Net GAN discriminator, which further improves image synthesis quality.
Image Segmentation with U-Net - Analytics Vidhya
Understanding Pix2Pix GAN. The name itself says “Pixel to Pixel
Understanding GAN Loss Functions
What is UNET? - Idiot Developer
218 - Difference between UpSampling2D and Conv2DTranspose used in
Applications of generative adversarial networks in neuroimaging
Image to Image Translation–U-Nets and cGANs - Josh Greaves
Medical image segmentation using deep learning: A survey - Wang
AI, Free Full-Text
U-Net Conditional GANs for Photo-Realistic and Identity-Preserving
Simplified schematic of U-net based GAN. The generator synthesizes
U-Net: Convolutional Networks for Biomedical Image Segmentation
Must-Read Papers on GANs. Generative Adversarial Networks are one