We propose to use a quality estimator and evolutionary methods to search the latent space of generative adversarial networks trained on small, difficult datasets, or both. The new method leads to the generation of significantly higher quality images while preserving the original generator's diversity. Human raters preferred an image from the new version with frequency 83.7pc for Cats, 74pc for FashionGen, 70.4pc for Horses, and 69.2pc for Artworks, and minor improvements for the already excellent GANs for faces. This approach applies to any quality scorer and GAN generator.
@article{arxiv.2009.13311,
title = {EvolGAN: Evolutionary Generative Adversarial Networks},
author = {Baptiste Roziere and Fabien Teytaud and Vlad Hosu and Hanhe Lin and Jeremy Rapin and Mariia Zameshina and Olivier Teytaud},
journal= {arXiv preprint arXiv:2009.13311},
year = {2020}
}