The GAN is dead; long live the GAN! A Modern GAN Baseline. Huang, Y., Gokaslan, A., Kuleshov, V., & Tompkin, J. January, 2025. arXiv:2501.05441 [cs]
The GAN is dead; long live the GAN! A Modern GAN Baseline [link]Paper  doi  abstract   bibtex   
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 Second, our new loss allows us to discard all ad-hoc tricks and replace outdated backbones used in common GANs with modern architectures.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.
@misc{huang_gan_2025,
	title = {The {GAN} is dead; long live the {GAN}! {A} {Modern} {GAN} {Baseline}},
	url = {http://arxiv.org/abs/2501.05441},
	doi = {10.48550/arXiv.2501.05441},
	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 Second, our new loss allows us to discard all ad-hoc tricks and replace outdated backbones used in common GANs with modern architectures.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.},
	urldate = {2025-01-15},
	publisher = {arXiv},
	author = {Huang, Yiwen and Gokaslan, Aaron and Kuleshov, Volodymyr and Tompkin, James},
	month = jan,
	year = {2025},
	note = {arXiv:2501.05441 [cs]},
	keywords = {\#ICML{\textgreater}24, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, ❤️},
}

Downloads: 0