Wasserstein Auto-Encoders

Ilya Tolstikhin, Olivier Bousquet, Sylvain Gelly, Bernhard Schoelkopf
11/5/2017
stat.MLcs.LG

Official code: tolstikhin/wae (511 stars, BSD-3-Clause, last commit 2018-06-28).

Abstract

We propose the Wasserstein Auto-Encoder (WAE)---a new algorithm for building a generative model of the data distribution. WAE minimizes a penalized form of the Wasserstein distance between the model distribution and the target distribution, which leads to a different regularizer than the one used by the Variational Auto-Encoder (VAE). This regularizer encourages the encoded training distribution to match the prior. We compare our algorithm with several other techniques and show that it is a generalization of adversarial auto-encoders (AAE). Our experiments show that WAE shares many of the properties of VAEs (stable training, encoder-decoder architecture, nice latent manifold structure) while generating samples of better quality, as measured by the FID score.

View on arXivView PDF

Code Implementations(1)

tolstikhin/waeOfficial100%
BSD-3-Clause

Wasserstein Auto-Encoders

51192PythonNov 10, 20178 years agoBSD-3-Clause

Similar Papers

Learning Autoencoders with Relational Regularization

88% match1 repo462020

Sliced-Wasserstein Autoencoder: An Embarrassingly Simple Generative Model

88% match1 repo922018

From Variational to Deterministic Autoencoders

87% match1 repo1262019

On the regularization of Wasserstein GANs

86% match1 repo92017

Analyzing and Improving Optimal-Transport-based Adversarial Networks

85% match2023

Perceptual Generative Autoencoders

85% match1 repo222019

Cite this paper

@article{tolstikhin2017wasserstein,
  title  = {Wasserstein Auto-Encoders},
  author = {Ilya Tolstikhin and Olivier Bousquet and Sylvain Gelly and Bernhard Schoelkopf},
  year   = {2017},
  eprint = {1711.01558},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1711.01558}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1711.01558.svg)](https://codeofpaper.com/paper/1711.01558)

Discussion