Our key ideais to leverage a trained GAN to extract pixel-wise represen-tation from the input image and use it as feature vectorsfor a segmentation network. - r/MachineLearning / pdf