This code was developed during my master's course at Lucerne University of Applied Sciences and Arts. The main part contains infrastructure to run different types of experiments around training GANs to generate images with different conditions. Furthermore, there are additional scripts to explore the trained models and analyze their learned representations.

It is also used for the paper "Applications of Generative Adversarial Networks to Dermatologic Imaging" by Fabian Furger, Ludovic Amruthalingam, Alexander Navarini and Marc Pouly, published in Workshop on Artificial Neural Networks in Pattern Recognition (ANNPR), 2020.

The framework is intended to be run inside a Docker container, where the specified python-packages are installed. Individual experiments can then be run with the corresponding makefile targets or by executing the relevant script in the src/ directory directly.

While the framework and experiments were developed with a particular target domain - different types of dermatology imaging - it can be applied to arbitrary image domains. Indeed, the same models have also been trained on other images, without any code changes or reconfiguration.