250 - Image to image translation using Pix2Pix GAN

A review of the original publication. https://arxiv.org/abs/1611.07004 Code generated in the video can be downloaded from here: https://github.com/bnsreenu/python_fo... The discriminator in the Pix2Pix GAN is implemented as a PatchGAN. PatchGAN discriminator tries to classify if each N×N patch in an image is real or fake. (as opposed to classifying an entire image) This discriminator is run convolutionally across the image, averaging all responses to provide the final output. The receptive field in a PatchGAN represents the relationship between one output activation to an area on the input image. A 70×70 PatchGAN will classify 70×70 patches of the input image as real or fake.