Implementing CycleGAN on Tensorflow
mage-to-image translation is the task of transforming an image from one domain (e.g., images of zebras), to another (e.g., images of horses). Ideally, other features of the image — anything not directly related to either domain, such as the background — should stay recognizably the same. As we might imagine, a good image-to-image translation system could have an almost unlimited number of applications. Changing art styles, going from sketch to photo, or changing the season of the landscape in a photo are just a few examples.
Read the actual blog here: https://towardsdatascience.com/cyclegan-learning-to-translate-images-without-paired-training-data-5b4e93862c8d
Read the paper here: https://arxiv.org/abs/1703.10593