Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities
Abstract
In this paper, we present the Lipschitz regularization theory and algorithms for a novel Loss-Sensitive Generative Adversarial Network (LS-GAN). Specifically, it trains a loss function to distinguish between real and fake samples by designated margins, while learning a generator alternately to produce realistic samples by minimizing their losses. The LS-GAN further regularizes its loss function with a Lipschitz regularity condition on the density of real data, yielding a regularized model that can better generalize to produce new data from a reasonable number of training examples than the classic GAN. We will further present a Generalized LS-GAN (GLS-GAN) and show it contains a large family of regularized GAN models, including both LS-GAN and Wasserstein GAN, as its special cases. Compared with the other GAN models, we will conduct experiments to show both LS-GAN and GLS-GAN exhibit competitive ability in generating new images in terms of the Minimum Reconstruction Error (MRE) assessed on a separate test set. We further extend the LS-GAN to a conditional form for supervised and semi-supervised learning problems, and demonstrate its outstanding performance on image classification tasks.
Cite
@article{arxiv.1701.06264,
title = {Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities},
author = {Guo-Jun Qi},
journal= {arXiv preprint arXiv:1701.06264},
year = {2019}
}
Comments
The source codes for both LS-GAN and GLS-GAN are available at \url{https://github.com/maple-research-lab}. LS-GAN is also supported by Microsoft CNTK at \url{https://www.cntk.ai/pythondocs/CNTK_206C_WGAN_LSGAN.html}. The original codes of LS-GAN and GLS-GAN are also available at https://github.com/guojunq/lsgan/ and https://github.com/guojunq/glsgan/