papers AI Learner
The Github is limit! Click to go to the new site.

The Numerics of GANs

2018-06-11
Lars Mescheder, Sebastian Nowozin, Andreas Geiger

Abstract

In this paper, we analyze the numerics of common algorithms for training Generative Adversarial Networks (GANs). Using the formalism of smooth two-player games we analyze the associated gradient vector field of GAN training objectives. Our findings suggest that the convergence of current algorithms suffers due to two factors: i) presence of eigenvalues of the Jacobian of the gradient vector field with zero real-part, and ii) eigenvalues with big imaginary part. Using these findings, we design a new algorithm that overcomes some of these limitations and has better convergence properties. Experimentally, we demonstrate its superiority on training common GAN architectures and show convergence on GAN architectures that are known to be notoriously hard to train.

Abstract (translated by Google)
URL

https://arxiv.org/abs/1705.10461

PDF

https://arxiv.org/pdf/1705.10461


Similar Posts

Comments