Delta-GAN-Encoder: Encoding Semantic Changes for Explicit Image Editing, using Few Synthetic Samples

Delta-GAN-Encoder is a project that introduces a novel method for learning to control any desired attribute in a pre-trained GAN's latent space, allowing for explicit image editing using only a few synthetic samples. The project addresses the complex task of understanding and controlling generative models' latent space, relying on minimal samples to achieve Sim2Real learning.