Thanks! Noise is actually preserved really well, and is one of the strengths of using a GAN. Check out this visual example from the "All Evaluation Images" link

Thanks! Noise is actually preserved really well, and is one of the strengths of using a GAN. Check out this visual example from the "All Evaluation Images" link: https://storage.googleapis.com/hific/clic2020/visualize.html?perPage=10&imgs=8c3f64635c07801bc2c9def490a27168 https://storage.googleapis.com/hific/clic2020/visualize.html...