Advanced Deep Learning with Python

Implementing CycleGAN

This example contains several source files located at https://github.com/PacktPublishing/Advanced-Deep-Learning-with-Python/tree/master/Chapter05/cyclegan. Besides TF, the code also depends on tensorflow_addons and imageio packages. You can install them with the pip package installer. We'll implement CycleGAN for multiple training datasets, all of which were provided by the authors of the paper. Before you run the example, you have to download the relevant dataset with the help of the download_dataset.sh executable script, which uses the dataset name as an argument. The list of available datasets is included in the file. Once you've downloaded this, you can access the images with the help of the DataLoader class, which ...