Comparing the interpretability of GAN architectures CSC413H1 Final Project - Winter 2021, University of Toronto (St. George) By Henry Tu & Seel Patel We compared the interpretability of StyleGAN and COCO-GAN using techniques such as C2ST, Image Sharpness, Color distribution, and Nearest Neighbours.