The GAN is dead; long live the GAN! A Modern GAN Baseline
Nick Huang, Aaron Gokaslan, Volodymyr Kuleshov, James Tompkin
Abstract
There is a widely-spread claim that GANs are difficult to train, and GAN architectures in the literature are littered with empirical tricks. We provide evidence against this claim and build a modern GAN baseline in a more principled manner. First, we derive a well-behaved regularized relativistic GAN loss that addresses issues of mode dropping and non-convergence that were previously tackled via a bag of ad-hoc tricks. We analyze our loss mathematically and prove that it admits local convergence guarantees, unlike most existing relativistic losses. Second, this loss allows us to discard all ad-hoc tricks and replace outdated backbones used in common GANs with modern architectures. Using StyleGAN2 as an example, we present a roadmap of simplification and modernization that results in a new minimalist baseline---R3GAN. Despite being simple, our approach surpasses StyleGAN2 on FFHQ, ImageNet, CIFAR, and Stacked MNIST datasets, and compares favorably against state-of-the-art GANs and diffusion models. Code: https://www.github.com/brownvc/R3GAN
BibTeX
@inproceedings{
huang2024the,
title={The {GAN} is dead; long live the {GAN}! A Modern {GAN} Baseline},
author={Nick Huang and Aaron Gokaslan and Volodymyr Kuleshov and James Tompkin},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=OrtN9hPP7V}
}