Fostering Diversity in Spatial Evolutionary Generative Adversarial Networks
Machine Learning
2021-06-28 v1 Artificial Intelligence
Distributed, Parallel, and Cluster Computing
Abstract
Generative adversary networks (GANs) suffer from training pathologies such as instability and mode collapse, which mainly arise from a lack of diversity in their adversarial interactions. Co-evolutionary GAN (CoE-GAN) training algorithms have shown to be resilient to these pathologies. This article introduces Mustangs, a spatially distributed CoE-GAN, which fosters diversity by using different loss functions during the training. Experimental analysis on MNIST and CelebA demonstrated that Mustangs trains statistically more accurate generators.
Cite
@article{arxiv.2106.13590,
title = {Fostering Diversity in Spatial Evolutionary Generative Adversarial Networks},
author = {Jamal Toutouh and Erik Hemberg and Una-May O'Reilly},
journal= {arXiv preprint arXiv:2106.13590},
year = {2021}
}
Comments
Accepted to be presented during Conference of the Spanish Association of Artificial Intelligence (CAEPIA 2021). arXiv admin note: substantial text overlap with arXiv:1905.12702