FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method

This paper proposes a fast optimization algorithm for the multichannel variational autoencoder (MVAE) method, a recently proposed powerful multichannel source separation technique. The MVAE method can achieve good source separation performance thanks to a convergence-guaranteed optimization algorith...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE access 2020, Vol.8, p.228740-228753
Hauptverfasser: Li, Li, Kameoka, Hirokazu, Inoue, Shota, Makino, Shoji
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 228753
container_issue
container_start_page 228740
container_title IEEE access
container_volume 8
creator Li, Li
Kameoka, Hirokazu
Inoue, Shota
Makino, Shoji
description This paper proposes a fast optimization algorithm for the multichannel variational autoencoder (MVAE) method, a recently proposed powerful multichannel source separation technique. The MVAE method can achieve good source separation performance thanks to a convergence-guaranteed optimization algorithm and the idea of jointly performing multi-speaker separation and speaker identification. However, one drawback is the high computational cost of the optimization algorithm. To overcome this drawback, this paper proposes using an auxiliary classifier VAE, an information-theoretic extension of the conditional VAE (CVAE), to train the generative model of the source spectrograms and using it to efficiently update the parameters of the source spectrogram models at each iteration of the source separation algorithm. We call the proposed algorithm "FastMVAE" (or fMVAE for short). Experimental evaluations revealed that the proposed fast algorithm can achieve high source separation performance in both speaker-dependent and speaker-independent scenarios while significantly reducing the computational time compared to the original MVAE method by more than 90% on both GPU and CPU. However, there is still room for improvement of about 3 dB compared to the original MVAE method.
doi_str_mv 10.1109/ACCESS.2020.3045704
format Article
fullrecord <record><control><sourceid>proquest_ieee_</sourceid><recordid>TN_cdi_ieee_primary_9298772</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9298772</ieee_id><doaj_id>oai_doaj_org_article_bb3cce8ae2e746a7ad257dbc2dba2cad</doaj_id><sourcerecordid>2474859728</sourcerecordid><originalsourceid>FETCH-LOGICAL-c544t-d859486e552f270fd33aaf39489862e18601909225220a24dbf5729def8fc41b3</originalsourceid><addsrcrecordid>eNpNUcFqGzEQXUoDDWm-IBdBz3a1I2kl9bYYpw3E5JAmlx6EVhrFMmvL1cqH9usrZ0PoXGZ4vPdmhtc0Ny1dti3VX_vVav34uAQKdMkoF5LyD80ltJ1eMMG6j__Nn5rradrRWqpCQl42v27tVDbP_fob6cl5Jg_HEvfxry0xHUg_vqQcy3ZPQsqkbJFsTmOJbmsPBxzJs83xlWhH0p9KwoNLHjPZYNkm_7m5CHac8PqtXzVPt-ufqx-L-4fvd6v-fuEE52XhldBcdSgEBJA0eMasDaxiWnWArepoq6kGEADUAvdDEBK0x6CC4-3Arpq72dcnuzPHHPc2_zHJRvMKpPxibK5Hj2iGgTmHyiKg5J2V1oOQfnDgBwvO-ur1ZfY65vT7hFMxu3TK9b_JAJe8nipBVRabWS6nacoY3re21JxDMXMo5hyKeQulqm5mVUTEd4UGraQE9g89_4gO</addsrcrecordid><sourcetype>Open Website</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2474859728</pqid></control><display><type>article</type><title>FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method</title><source>IEEE Open Access Journals</source><source>DOAJ Directory of Open Access Journals</source><source>Elektronische Zeitschriftenbibliothek - Frei zugängliche E-Journals</source><creator>Li, Li ; Kameoka, Hirokazu ; Inoue, Shota ; Makino, Shoji</creator><creatorcontrib>Li, Li ; Kameoka, Hirokazu ; Inoue, Shota ; Makino, Shoji</creatorcontrib><description>This paper proposes a fast optimization algorithm for the multichannel variational autoencoder (MVAE) method, a recently proposed powerful multichannel source separation technique. The MVAE method can achieve good source separation performance thanks to a convergence-guaranteed optimization algorithm and the idea of jointly performing multi-speaker separation and speaker identification. However, one drawback is the high computational cost of the optimization algorithm. To overcome this drawback, this paper proposes using an auxiliary classifier VAE, an information-theoretic extension of the conditional VAE (CVAE), to train the generative model of the source spectrograms and using it to efficiently update the parameters of the source spectrogram models at each iteration of the source separation algorithm. We call the proposed algorithm "FastMVAE" (or fMVAE for short). Experimental evaluations revealed that the proposed fast algorithm can achieve high source separation performance in both speaker-dependent and speaker-independent scenarios while significantly reducing the computational time compared to the original MVAE method by more than 90% on both GPU and CPU. However, there is still room for improvement of about 3 dB compared to the original MVAE method.</description><identifier>ISSN: 2169-3536</identifier><identifier>EISSN: 2169-3536</identifier><identifier>DOI: 10.1109/ACCESS.2020.3045704</identifier><identifier>CODEN: IAECCG</identifier><language>eng</language><publisher>Piscataway: IEEE</publisher><subject>Algorithms ; auxiliary classifier VAE ; Computational efficiency ; Computational modeling ; Computing costs ; Computing time ; Decoding ; FastMVAE algorithm ; Information theory ; Iterative methods ; Multichannel source separation ; multichannel variational autoencoder (MVAE) method ; Neural networks ; Optimization ; Optimization algorithms ; Separation ; Source separation ; Spectrogram ; Spectrograms ; Task analysis</subject><ispartof>IEEE access, 2020, Vol.8, p.228740-228753</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c544t-d859486e552f270fd33aaf39489862e18601909225220a24dbf5729def8fc41b3</citedby><cites>FETCH-LOGICAL-c544t-d859486e552f270fd33aaf39489862e18601909225220a24dbf5729def8fc41b3</cites><orcidid>0000-0003-1934-640X ; 0000-0003-3102-0162 ; 0000-0002-3121-7857</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9298772$$EHTML$$P50$$Gieee$$Hfree_for_read</linktohtml><link.rule.ids>314,780,784,864,2100,4022,27632,27922,27923,27924,54932</link.rule.ids></links><search><creatorcontrib>Li, Li</creatorcontrib><creatorcontrib>Kameoka, Hirokazu</creatorcontrib><creatorcontrib>Inoue, Shota</creatorcontrib><creatorcontrib>Makino, Shoji</creatorcontrib><title>FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method</title><title>IEEE access</title><addtitle>Access</addtitle><description>This paper proposes a fast optimization algorithm for the multichannel variational autoencoder (MVAE) method, a recently proposed powerful multichannel source separation technique. The MVAE method can achieve good source separation performance thanks to a convergence-guaranteed optimization algorithm and the idea of jointly performing multi-speaker separation and speaker identification. However, one drawback is the high computational cost of the optimization algorithm. To overcome this drawback, this paper proposes using an auxiliary classifier VAE, an information-theoretic extension of the conditional VAE (CVAE), to train the generative model of the source spectrograms and using it to efficiently update the parameters of the source spectrogram models at each iteration of the source separation algorithm. We call the proposed algorithm "FastMVAE" (or fMVAE for short). Experimental evaluations revealed that the proposed fast algorithm can achieve high source separation performance in both speaker-dependent and speaker-independent scenarios while significantly reducing the computational time compared to the original MVAE method by more than 90% on both GPU and CPU. However, there is still room for improvement of about 3 dB compared to the original MVAE method.</description><subject>Algorithms</subject><subject>auxiliary classifier VAE</subject><subject>Computational efficiency</subject><subject>Computational modeling</subject><subject>Computing costs</subject><subject>Computing time</subject><subject>Decoding</subject><subject>FastMVAE algorithm</subject><subject>Information theory</subject><subject>Iterative methods</subject><subject>Multichannel source separation</subject><subject>multichannel variational autoencoder (MVAE) method</subject><subject>Neural networks</subject><subject>Optimization</subject><subject>Optimization algorithms</subject><subject>Separation</subject><subject>Source separation</subject><subject>Spectrogram</subject><subject>Spectrograms</subject><subject>Task analysis</subject><issn>2169-3536</issn><issn>2169-3536</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>ESBDL</sourceid><sourceid>RIE</sourceid><sourceid>DOA</sourceid><recordid>eNpNUcFqGzEQXUoDDWm-IBdBz3a1I2kl9bYYpw3E5JAmlx6EVhrFMmvL1cqH9usrZ0PoXGZ4vPdmhtc0Ny1dti3VX_vVav34uAQKdMkoF5LyD80ltJ1eMMG6j__Nn5rradrRWqpCQl42v27tVDbP_fob6cl5Jg_HEvfxry0xHUg_vqQcy3ZPQsqkbJFsTmOJbmsPBxzJs83xlWhH0p9KwoNLHjPZYNkm_7m5CHac8PqtXzVPt-ufqx-L-4fvd6v-fuEE52XhldBcdSgEBJA0eMasDaxiWnWArepoq6kGEADUAvdDEBK0x6CC4-3Arpq72dcnuzPHHPc2_zHJRvMKpPxibK5Hj2iGgTmHyiKg5J2V1oOQfnDgBwvO-ur1ZfY65vT7hFMxu3TK9b_JAJe8nipBVRabWS6nacoY3re21JxDMXMo5hyKeQulqm5mVUTEd4UGraQE9g89_4gO</recordid><startdate>2020</startdate><enddate>2020</enddate><creator>Li, Li</creator><creator>Kameoka, Hirokazu</creator><creator>Inoue, Shota</creator><creator>Makino, Shoji</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>ESBDL</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>7SR</scope><scope>8BQ</scope><scope>8FD</scope><scope>JG9</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>DOA</scope><orcidid>https://orcid.org/0000-0003-1934-640X</orcidid><orcidid>https://orcid.org/0000-0003-3102-0162</orcidid><orcidid>https://orcid.org/0000-0002-3121-7857</orcidid></search><sort><creationdate>2020</creationdate><title>FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method</title><author>Li, Li ; Kameoka, Hirokazu ; Inoue, Shota ; Makino, Shoji</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c544t-d859486e552f270fd33aaf39489862e18601909225220a24dbf5729def8fc41b3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Algorithms</topic><topic>auxiliary classifier VAE</topic><topic>Computational efficiency</topic><topic>Computational modeling</topic><topic>Computing costs</topic><topic>Computing time</topic><topic>Decoding</topic><topic>FastMVAE algorithm</topic><topic>Information theory</topic><topic>Iterative methods</topic><topic>Multichannel source separation</topic><topic>multichannel variational autoencoder (MVAE) method</topic><topic>Neural networks</topic><topic>Optimization</topic><topic>Optimization algorithms</topic><topic>Separation</topic><topic>Source separation</topic><topic>Spectrogram</topic><topic>Spectrograms</topic><topic>Task analysis</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Li, Li</creatorcontrib><creatorcontrib>Kameoka, Hirokazu</creatorcontrib><creatorcontrib>Inoue, Shota</creatorcontrib><creatorcontrib>Makino, Shoji</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE Open Access Journals</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics &amp; Communications Abstracts</collection><collection>Engineered Materials Abstracts</collection><collection>METADEX</collection><collection>Technology Research Database</collection><collection>Materials Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts – Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>DOAJ Directory of Open Access Journals</collection><jtitle>IEEE access</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Li, Li</au><au>Kameoka, Hirokazu</au><au>Inoue, Shota</au><au>Makino, Shoji</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method</atitle><jtitle>IEEE access</jtitle><stitle>Access</stitle><date>2020</date><risdate>2020</risdate><volume>8</volume><spage>228740</spage><epage>228753</epage><pages>228740-228753</pages><issn>2169-3536</issn><eissn>2169-3536</eissn><coden>IAECCG</coden><abstract>This paper proposes a fast optimization algorithm for the multichannel variational autoencoder (MVAE) method, a recently proposed powerful multichannel source separation technique. The MVAE method can achieve good source separation performance thanks to a convergence-guaranteed optimization algorithm and the idea of jointly performing multi-speaker separation and speaker identification. However, one drawback is the high computational cost of the optimization algorithm. To overcome this drawback, this paper proposes using an auxiliary classifier VAE, an information-theoretic extension of the conditional VAE (CVAE), to train the generative model of the source spectrograms and using it to efficiently update the parameters of the source spectrogram models at each iteration of the source separation algorithm. We call the proposed algorithm "FastMVAE" (or fMVAE for short). Experimental evaluations revealed that the proposed fast algorithm can achieve high source separation performance in both speaker-dependent and speaker-independent scenarios while significantly reducing the computational time compared to the original MVAE method by more than 90% on both GPU and CPU. However, there is still room for improvement of about 3 dB compared to the original MVAE method.</abstract><cop>Piscataway</cop><pub>IEEE</pub><doi>10.1109/ACCESS.2020.3045704</doi><tpages>14</tpages><orcidid>https://orcid.org/0000-0003-1934-640X</orcidid><orcidid>https://orcid.org/0000-0003-3102-0162</orcidid><orcidid>https://orcid.org/0000-0002-3121-7857</orcidid><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 2169-3536
ispartof IEEE access, 2020, Vol.8, p.228740-228753
issn 2169-3536
2169-3536
language eng
recordid cdi_ieee_primary_9298772
source IEEE Open Access Journals; DOAJ Directory of Open Access Journals; Elektronische Zeitschriftenbibliothek - Frei zugängliche E-Journals
subjects Algorithms
auxiliary classifier VAE
Computational efficiency
Computational modeling
Computing costs
Computing time
Decoding
FastMVAE algorithm
Information theory
Iterative methods
Multichannel source separation
multichannel variational autoencoder (MVAE) method
Neural networks
Optimization
Optimization algorithms
Separation
Source separation
Spectrogram
Spectrograms
Task analysis
title FastMVAE: A Fast Optimization Algorithm for the Multichannel Variational Autoencoder Method
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-12T12%3A23%3A00IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_ieee_&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=FastMVAE:%20A%20Fast%20Optimization%20Algorithm%20for%20the%20Multichannel%20Variational%20Autoencoder%20Method&rft.jtitle=IEEE%20access&rft.au=Li,%20Li&rft.date=2020&rft.volume=8&rft.spage=228740&rft.epage=228753&rft.pages=228740-228753&rft.issn=2169-3536&rft.eissn=2169-3536&rft.coden=IAECCG&rft_id=info:doi/10.1109/ACCESS.2020.3045704&rft_dat=%3Cproquest_ieee_%3E2474859728%3C/proquest_ieee_%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2474859728&rft_id=info:pmid/&rft_ieee_id=9298772&rft_doaj_id=oai_doaj_org_article_bb3cce8ae2e746a7ad257dbc2dba2cad&rfr_iscdi=true