Hey Tomer, me and Bar Firuan would like to explore the mentioned subject for the course project. it is a simple technique to analyze Generative Adversarial Networks (GANs) and create interpretable controls for image synthesis, such as change of viewpoint, aging, lighting, and time of day. Thanks! GitHub - harskish/ganspace: Discovering Interpretable GAN Controls [NeurIPS 2020]
Project approved.