Cramer-Wold AutoEncoder

Cramer-Wold AutoEncoder

We propose a new generative model, Cramer-Wold Autoencoder (CWAE). Following WAE, we directly encourage normality of the latent space. Our paper uses also the recent idea from Sliced WAE (SWAE) model, which uses one-dimensional projections as a method of verifying closeness of two distributions. The...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2019-07
Hauptverfasser: Knop, Szymon, Tabor, Jacek, Spurek, Przemysław, Podolak, Igor, Mazur, Marcin, Jastrzębski, Stanisław
Format: Artikel
Sprache:eng
Schlagworte:
Normality
Optimization
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Knop, Szymon
Tabor, Jacek
Spurek, Przemysław
Podolak, Igor
Mazur, Marcin
Jastrzębski, Stanisław
description We propose a new generative model, Cramer-Wold Autoencoder (CWAE). Following WAE, we directly encourage normality of the latent space. Our paper uses also the recent idea from Sliced WAE (SWAE) model, which uses one-dimensional projections as a method of verifying closeness of two distributions. The crucial new ingredient is the introduction of a new (Cramer-Wold) metric in the space of densities, which replaces the Wasserstein metric used in SWAE. We show that the Cramer-Wold metric between Gaussian mixtures is given by a simple analytic formula, which results in the removal of sampling necessary to estimate the cost function in WAE and SWAE models. As a consequence, while drastically simplifying the optimization procedure, CWAE produces samples of a matching perceptual quality to other SOTA models.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2073435511</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2073435511</sourcerecordid><originalsourceid>FETCH-proquest_journals_20734355113</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQQdy5KzE0t0g3Pz0lRcCwtyXfNS85PSS3iYWBNS8wpTuWF0twMym6uIc4eugVF-YWlqcUl8Vn5pUV5QKl4IwNzYxNjU1NDQ2PiVAEAGDEpcA</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2073435511</pqid></control><display><type>article</type><title>Cramer-Wold AutoEncoder</title><source>Free E- Journals</source><creator>Knop, Szymon ; Tabor, Jacek ; Spurek, Przemysław ; Podolak, Igor ; Mazur, Marcin ; Jastrzębski, Stanisław</creator><creatorcontrib>Knop, Szymon ; Tabor, Jacek ; Spurek, Przemysław ; Podolak, Igor ; Mazur, Marcin ; Jastrzębski, Stanisław</creatorcontrib><description>We propose a new generative model, Cramer-Wold Autoencoder (CWAE). Following WAE, we directly encourage normality of the latent space. Our paper uses also the recent idea from Sliced WAE (SWAE) model, which uses one-dimensional projections as a method of verifying closeness of two distributions. The crucial new ingredient is the introduction of a new (Cramer-Wold) metric in the space of densities, which replaces the Wasserstein metric used in SWAE. We show that the Cramer-Wold metric between Gaussian mixtures is given by a simple analytic formula, which results in the removal of sampling necessary to estimate the cost function in WAE and SWAE models. As a consequence, while drastically simplifying the optimization procedure, CWAE produces samples of a matching perceptual quality to other SOTA models.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Normality ; Optimization</subject><ispartof>arXiv.org, 2019-07</ispartof><rights>2019. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>780,784</link.rule.ids></links><search><creatorcontrib>Knop, Szymon</creatorcontrib><creatorcontrib>Tabor, Jacek</creatorcontrib><creatorcontrib>Spurek, Przemysław</creatorcontrib><creatorcontrib>Podolak, Igor</creatorcontrib><creatorcontrib>Mazur, Marcin</creatorcontrib><creatorcontrib>Jastrzębski, Stanisław</creatorcontrib><title>Cramer-Wold AutoEncoder</title><title>arXiv.org</title><description>We propose a new generative model, Cramer-Wold Autoencoder (CWAE). Following WAE, we directly encourage normality of the latent space. Our paper uses also the recent idea from Sliced WAE (SWAE) model, which uses one-dimensional projections as a method of verifying closeness of two distributions. The crucial new ingredient is the introduction of a new (Cramer-Wold) metric in the space of densities, which replaces the Wasserstein metric used in SWAE. We show that the Cramer-Wold metric between Gaussian mixtures is given by a simple analytic formula, which results in the removal of sampling necessary to estimate the cost function in WAE and SWAE models. As a consequence, while drastically simplifying the optimization procedure, CWAE produces samples of a matching perceptual quality to other SOTA models.</description><subject>Normality</subject><subject>Optimization</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2019</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQQdy5KzE0t0g3Pz0lRcCwtyXfNS85PSS3iYWBNS8wpTuWF0twMym6uIc4eugVF-YWlqcUl8Vn5pUV5QKl4IwNzYxNjU1NDQ2PiVAEAGDEpcA</recordid><startdate>20190702</startdate><enddate>20190702</enddate><creator>Knop, Szymon</creator><creator>Tabor, Jacek</creator><creator>Spurek, Przemysław</creator><creator>Podolak, Igor</creator><creator>Mazur, Marcin</creator><creator>Jastrzębski, Stanisław</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20190702</creationdate><title>Cramer-Wold AutoEncoder</title><author>Knop, Szymon ; Tabor, Jacek ; Spurek, Przemysław ; Podolak, Igor ; Mazur, Marcin ; Jastrzębski, Stanisław</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_20734355113</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2019</creationdate><topic>Normality</topic><topic>Optimization</topic><toplevel>online_resources</toplevel><creatorcontrib>Knop, Szymon</creatorcontrib><creatorcontrib>Tabor, Jacek</creatorcontrib><creatorcontrib>Spurek, Przemysław</creatorcontrib><creatorcontrib>Podolak, Igor</creatorcontrib><creatorcontrib>Mazur, Marcin</creatorcontrib><creatorcontrib>Jastrzębski, Stanisław</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Knop, Szymon</au><au>Tabor, Jacek</au><au>Spurek, Przemysław</au><au>Podolak, Igor</au><au>Mazur, Marcin</au><au>Jastrzębski, Stanisław</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Cramer-Wold AutoEncoder</atitle><jtitle>arXiv.org</jtitle><date>2019-07-02</date><risdate>2019</risdate><eissn>2331-8422</eissn><abstract>We propose a new generative model, Cramer-Wold Autoencoder (CWAE). Following WAE, we directly encourage normality of the latent space. Our paper uses also the recent idea from Sliced WAE (SWAE) model, which uses one-dimensional projections as a method of verifying closeness of two distributions. The crucial new ingredient is the introduction of a new (Cramer-Wold) metric in the space of densities, which replaces the Wasserstein metric used in SWAE. We show that the Cramer-Wold metric between Gaussian mixtures is given by a simple analytic formula, which results in the removal of sampling necessary to estimate the cost function in WAE and SWAE models. As a consequence, while drastically simplifying the optimization procedure, CWAE produces samples of a matching perceptual quality to other SOTA models.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2019-07
issn 2331-8422
language eng
recordid cdi_proquest_journals_2073435511
source Free E- Journals
subjects Normality
Optimization
title Cramer-Wold AutoEncoder
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-05T20%3A24%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Cramer-Wold%20AutoEncoder&rft.jtitle=arXiv.org&rft.au=Knop,%20Szymon&rft.date=2019-07-02&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2073435511%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2073435511&rft_id=info:pmid/&rfr_iscdi=true