no code implementations • ICLR 2019 • Chad Atalla, Bartholomew Tam, Amanda Song, Gary Cottrell
Existing methods of image to image translation require multiple steps in the training or modification process, and suffer from either an inability to generalize, or long training times.