We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Statistics > Machine Learning

Title: Wasserstein Auto-Encoders

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.
Comments: Published at ICLR 2018.. Included much wider hyperparameter sweep: in significant improvements in FIDs on CelebA
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Cite as: arXiv:1711.01558 [stat.ML]
  (or arXiv:1711.01558v4 [stat.ML] for this version)

Submission history

From: Ilya Tolstikhin [view email]
[v1] Sun, 5 Nov 2017 10:18:27 GMT (4254kb,D)
[v2] Tue, 19 Dec 2017 08:24:54 GMT (4254kb,D)
[v3] Mon, 12 Mar 2018 15:06:34 GMT (4262kb,D)
[v4] Thu, 5 Dec 2019 10:27:44 GMT (4307kb,D)

Link back to: arXiv, form interface, contact.