Generative Modeling using the Sliced Wasserstein Distance

03/29/2018
by   Ishan Deshpande, et al.
2

Generative Adversarial Nets (GANs) are very successful at modeling distributions from given samples, even in the high-dimensional case. However, their formulation is also known to be hard to optimize and often not stable. While this is particularly true for early GAN formulations, there has been significant empirically motivated and theoretically founded progress to improve stability, for instance, by using the Wasserstein distance rather than the Jenson-Shannon divergence. Here, we consider an alternative formulation for generative modeling based on random projections which, in its simplest form, results in a single objective rather than a saddle-point formulation. By augmenting this approach with a discriminator we improve its accuracy. We found our approach to be significantly more stable compared to even the improved Wasserstein GAN. Further, unlike the traditional GAN loss, the loss formulated in our method is a good measure of the actual distance between the distributions and, for the first time for GAN training, we are able to show estimates for the same.

READ FULL TEXT

Authors

page 1

page 6

page 7

page 12

04/18/2019

From GAN to WGAN

This paper explains the math behind a generative adversarial network (GA...
06/19/2017

Dualing GANs

Generative adversarial nets (GANs) are a promising technique for modelin...
10/28/2018

A Convex Duality Framework for GANs

Generative adversarial network (GAN) is a minimax game between a generat...
04/10/2019

Sliced Wasserstein Generative Models

In generative modeling, the Wasserstein distance (WD) has emerged as a u...
04/11/2019

Max-Sliced Wasserstein Distance and its use for GANs

Generative adversarial nets (GANs) and variational auto-encoders have si...
01/26/2017

Wasserstein GAN

We introduce a new algorithm named WGAN, an alternative to traditional G...
11/24/2017

Wasserstein Introspective Neural Networks

We present Wasserstein introspective neural networks (WINN) that are bot...

Code Repositories

CVProject-IAN

IAN for neural photo editing


view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.