Wasserstein GAN
Year: 2017
Type: preprint
Abstract: We introduce a new algorithm named WGAN, an alternative to traditional GANtraining. In this new model, we show that we can improve the stability of
learning, get rid of problems like mode collapse, and provide meaningful
learning curves useful for debugging and hyperparameter searches. Furthermore,
we show that the corresponding optimization problem is sound, and provide
extensive theoretical work highlighting the deep connections to other distances
between distributions.
Source: arXiv (Cornell University)
Authors: Martín Arjovsky, Soumith Chintala, Léon Bottou
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3037567775