Majorization-Minimization Algorithm for Discriminative Non-Negative Matrix Factorization

This paper proposes a basis training algorithm for discriminative non-negative matrix factorization (NMF) with applications to single-channel audio source separation. With an NMF-based approach to supervised audio source separation, NMF is first applied to train the basis spectra of each source usin...

Full description

Saved in:

Bibliographic Details
Published in:	IEEE access Vol. 8; pp. 227399 - 227408
Main Authors:	Li, Li, Kameoka, Hirokazu, Makino, Shoji
Format:	Journal Article
Language:	English
Published:	Piscataway IEEE 2020 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:	Algorithms Convergence Discriminative non-negative matrix factorization (NMF) Factorization Linear programming majorization-minimization Optimization Separation Signal processing algorithms single-channel signal processing Source separation Spectra Spectrogram speech enhancement Testing time Training Wiener filtering
ISSN:	2169-3536, 2169-3536
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Abstract	This paper proposes a basis training algorithm for discriminative non-negative matrix factorization (NMF) with applications to single-channel audio source separation. With an NMF-based approach to supervised audio source separation, NMF is first applied to train the basis spectra of each source using training examples and then applied to the spectrogram of a mixture signal using the pretrained basis spectra at test time. The source signals can then be separated out using a Wiener filter. Here, a typical way to train the basis spectra is to minimize the dissimilarity measure between the observed spectrogram and the NMF model. However, obtaining the basis spectra in this way does not ensure that the separated signal will be optimal at test time due to the inconsistency between the objective functions for training and separation (Wiener filtering). To address this mismatch, a framework called discriminative NMF (DNMF) has recently been proposed. While this framework is noteworthy in that it uses a common objective function for training and separation, the objective function becomes more analytically complex than that of regular NMF. In the original DNMF work, a multiplicative update algorithm was proposed for the basis training; however, the convergence of the algorithm is not guaranteed and can be very slow. To overcome this weakness, this paper proposes a convergence-guaranteed algorithm for DNMF based on a majorization-minimization principle. Experimental results show that the proposed algorithm outperform the conventional DNMF algorithm as well as the regular NMF algorithm in terms of both the signal-to-distortion and signal-to-interference ratios.
AbstractList	This paper proposes a basis training algorithm for discriminative non-negative matrix factorization (NMF) with applications to single-channel audio source separation. With an NMF-based approach to supervised audio source separation, NMF is first applied to train the basis spectra of each source using training examples and then applied to the spectrogram of a mixture signal using the pretrained basis spectra at test time. The source signals can then be separated out using a Wiener filter. Here, a typical way to train the basis spectra is to minimize the dissimilarity measure between the observed spectrogram and the NMF model. However, obtaining the basis spectra in this way does not ensure that the separated signal will be optimal at test time due to the inconsistency between the objective functions for training and separation (Wiener filtering). To address this mismatch, a framework called discriminative NMF (DNMF) has recently been proposed. While this framework is noteworthy in that it uses a common objective function for training and separation, the objective function becomes more analytically complex than that of regular NMF. In the original DNMF work, a multiplicative update algorithm was proposed for the basis training; however, the convergence of the algorithm is not guaranteed and can be very slow. To overcome this weakness, this paper proposes a convergence-guaranteed algorithm for DNMF based on a majorization-minimization principle. Experimental results show that the proposed algorithm outperform the conventional DNMF algorithm as well as the regular NMF algorithm in terms of both the signal-to-distortion and signal-to-interference ratios.
Author	Li, Li Makino, Shoji Kameoka, Hirokazu
Author_xml	– sequence: 1 givenname: Li orcidid: 0000-0002-3121-7857 surname: Li fullname: Li, Li email: lili@mmlab.cs.tsukuba.ac.jp organization: Graduate School of Systems and Information Engineering, University of Tsukuba, Ibaraki, Japan – sequence: 2 givenname: Hirokazu orcidid: 0000-0003-3102-0162 surname: Kameoka fullname: Kameoka, Hirokazu organization: NTT Communication Science Laboratories, Kanagawa, Japan – sequence: 3 givenname: Shoji orcidid: 0000-0003-1934-640X surname: Makino fullname: Makino, Shoji organization: Graduate School of Systems and Information Engineering, University of Tsukuba, Ibaraki, Japan
BookMark	eNqFUctOwzAQtFCReH4Bl0icU-ysE8fHqrRQicIBkLhZG8cprkoMjouAr8clVYW44Mt6HzOz9hyRQetaQ8gZo0PGqLwYjceT-_thRjM6BMpzIdkeOcxYIVPIoRj8uh-Q065b0njKWMrFIXma49J5-4XBujad29a-bJNktFrETnh-SRrnk0vbaR-bbWy-m-Q2Tt-aRZ_MMXj7kUxRhx3XCdlvcNWZ0208Jo_TycP4Or25u5qNRzepzjkPaY6c1o0oNYOSSagNa8RmPSjyuuCVqOOLaAWM0rzKRAOABjiDoqohazgiHJNZz1s7XKrXuCL6T-XQqp-C8wuFPli9MspQHsWAaY0ZL6VALhpNmaY1rZAXELnOe65X797Wpgtq6da-jeurjAte5pIDj1PQT2nvus6bZqfKqNo4onpH1MYRtXUkouQflLbh56eCR7v6B3vWY60xZqcmM1kKQeEbcZua4g
CODEN	IAECCG
CitedBy_id	crossref_primary_10_3390_info14070352
Cites_doi	10.1109/ICASSP.2014.6854302 10.1109/TASLP.2014.2320575 10.1109/ICASSP.2017.7952154 10.1587/transinf.2015EDL8114 10.1109/ICASSP.2014.6853860 10.1109/HSCMA.2014.6843241 10.1109/ISCAS.1999.777511 10.21437/Interspeech.2013-232 10.1109/HSCMA.2017.7895578 10.1111/j.2517-6161.1977.tb01600.x 10.1162/neco.2008.04-08-771 10.1007/978-3-319-53547-0_31 10.1007/978-4-431-55387-8_2 10.1109/TSA.2005.858005 10.1109/ICASSP.2016.7471631
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020
DBID	97E ESBDL RIA RIE AAYXX CITATION 7SC 7SP 7SR 8BQ 8FD JG9 JQ2 L7M L~C L~D DOA
DOI	10.1109/ACCESS.2020.3045791
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005-present IEEE Open Access Journals IEEE All-Society Periodicals Package (ASPP) 1998-Present IEEE Electronic Library (IEL) CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Engineered Materials Abstracts METADEX Technology Research Database Materials Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional DOAJ Directory of Open Access Journals
DatabaseTitle	CrossRef Materials Research Database Engineered Materials Abstracts Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace METADEX Computer and Information Systems Abstracts Professional
DatabaseTitleList	Materials Research Database
Database_xml	– sequence: 1 dbid: DOA name: DOAJ Directory of Open Access Journals url: https://www.doaj.org/ sourceTypes: Open Website – sequence: 2 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	2169-3536
EndPage	227408
ExternalDocumentID	oai_doaj_org_article_e0440d31cca24897a47fc01c0d0ba463 10_1109_ACCESS_2020_3045791 9298770
Genre	orig-research
GrantInformation_xml	– fundername: JSPS KAKENHI grantid: 18J20059 funderid: 10.13039/501100001691 – fundername: SECOM Science and Technology Foundation funderid: 10.13039/501100004298 – fundername: JST CREST grantid: JPMJCR19A3 funderid: 10.13039/501100003382
GroupedDBID	0R~ 4.4 5VS 6IK 97E AAJGR ABAZT ABVLG ACGFS ADBBV AGSQL ALMA_UNASSIGNED_HOLDINGS BCNDV BEFXN BFFAM BGNUA BKEBE BPEOZ EBS EJD ESBDL GROUPED_DOAJ IPLJI JAVBF KQ8 M43 M~E O9- OCL OK1 RIA RIE RNS AAYXX CITATION 7SC 7SP 7SR 8BQ 8FD JG9 JQ2 L7M L~C L~D
ID	FETCH-LOGICAL-c544t-5a40df78c138193de1f70008365d64b7d3040b31005b27f33ae34136bd32f4aa3
IEDL.DBID	DOA
ISICitedReferencesCount	1
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000604520500001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	2169-3536
IngestDate	Fri Oct 03 12:30:07 EDT 2025 Mon Jun 30 05:42:32 EDT 2025 Sat Nov 29 06:11:47 EST 2025 Tue Nov 18 21:08:02 EST 2025 Wed Aug 27 02:29:13 EDT 2025
IsDoiOpenAccess	true
IsOpenAccess	true
IsPeerReviewed	true
IsScholarly	true
Language	English
License	https://creativecommons.org/licenses/by/4.0/legalcode
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c544t-5a40df78c138193de1f70008365d64b7d3040b31005b27f33ae34136bd32f4aa3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0003-1934-640X 0000-0003-3102-0162 0000-0002-3121-7857
OpenAccessLink	https://doaj.org/article/e0440d31cca24897a47fc01c0d0ba463
PQID	2474859434
PQPubID	4845423
PageCount	10
ParticipantIDs	crossref_primary_10_1109_ACCESS_2020_3045791 ieee_primary_9298770 proquest_journals_2474859434 crossref_citationtrail_10_1109_ACCESS_2020_3045791 doaj_primary_oai_doaj_org_article_e0440d31cca24897a47fc01c0d0ba463
PublicationCentury	2000
PublicationDate	20200000 2020-00-00 20200101 2020-01-01
PublicationDateYYYYMMDD	2020-01-01
PublicationDate_xml	– year: 2020 text: 20200000
PublicationDecade	2020
PublicationPlace	Piscataway
PublicationPlace_xml	– name: Piscataway
PublicationTitle	IEEE access
PublicationTitleAbbrev	Access
PublicationYear	2020
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref13 garofolo (ref22) 1993 ref15 ref1 ref17 ref16 ref19 ref18 roweis (ref3) 2001 nakano (ref20) 2010 bach (ref5) 2005 lee (ref10) 2001 smaragdis (ref6) 2007 ref24 weninger (ref11) 2014 ref25 ref21 vincent (ref23) 2020 boulanger-lewandowski (ref14) 2012 ref8 ref7 ref9 ref4 grais (ref12) 2013 wang (ref2) 2006
References_xml	– ident: ref15 doi: 10.1109/ICASSP.2014.6854302 – year: 2020 ident: ref23 publication-title: The 4th CHiME speech separation and recognition challenge – ident: ref13 doi: 10.1109/TASLP.2014.2320575 – ident: ref9 doi: 10.1109/ICASSP.2017.7952154 – start-page: 865 year: 2014 ident: ref11 article-title: Discriminative NMF and its application to single-channel source separation publication-title: Proc INTERSPEECH – ident: ref16 doi: 10.1587/transinf.2015EDL8114 – start-page: 414 year: 2007 ident: ref6 article-title: Supervised and semisupervised separation of sounds from single-channel mixtures publication-title: Proc Integr Comput -Aided Eng – start-page: 793 year: 2001 ident: ref3 article-title: One microphone source separation publication-title: Proc NIPS – ident: ref7 doi: 10.1109/ICASSP.2014.6853860 – start-page: 65 year: 2005 ident: ref5 article-title: Blind one-microphone speech separation: A spectral learning approach publication-title: Proc NIPS – ident: ref17 doi: 10.1109/HSCMA.2014.6843241 – ident: ref4 doi: 10.1109/ISCAS.1999.777511 – start-page: 808 year: 2013 ident: ref12 article-title: Discriminative nonnegative dictionary learning using cross-coherence penalties for single channel source separation publication-title: Proc INTERSPEECH doi: 10.21437/Interspeech.2013-232 – ident: ref1 doi: 10.1109/HSCMA.2017.7895578 – start-page: 556 year: 2001 ident: ref10 article-title: Algorithms for non-negative matrix factorization publication-title: Proc NIPS – ident: ref19 doi: 10.1111/j.2517-6161.1977.tb01600.x – ident: ref18 doi: 10.1162/neco.2008.04-08-771 – year: 2006 ident: ref2 publication-title: Computational Auditory Scene Analysis Principles Algorithms and Applications – start-page: 283 year: 2010 ident: ref20 article-title: Convergence-guaranteed multiplicative algorithms for nonnegative matrix factorization with beta-divergence publication-title: Proc MLSP – year: 1993 ident: ref22 publication-title: Csr-i (wsj0) complete ldc93s6a' Web download – ident: ref25 doi: 10.1007/978-3-319-53547-0_31 – ident: ref21 doi: 10.1007/978-4-431-55387-8_2 – ident: ref24 doi: 10.1109/TSA.2005.858005 – start-page: 205 year: 2012 ident: ref14 article-title: Discriminative non-negative matrix factorization for multiple pitch estimation publication-title: Proc ISMIR – ident: ref8 doi: 10.1109/ICASSP.2016.7471631
SSID	ssj0000816957
Score	2.1679509
Snippet	This paper proposes a basis training algorithm for discriminative non-negative matrix factorization (NMF) with applications to single-channel audio source...
SourceID	doaj proquest crossref ieee
SourceType	Open Website Aggregation Database Enrichment Source Index Database Publisher
StartPage	227399
SubjectTerms	Algorithms Convergence Discriminative non-negative matrix factorization (NMF) Factorization Linear programming majorization-minimization Optimization Separation Signal processing algorithms single-channel signal processing Source separation Spectra Spectrogram speech enhancement Testing time Training Wiener filtering
SummonAdditionalLinks	– databaseName: IEEE Electronic Library (IEL) dbid: RIE link: http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3NT8MgFH9R40EPfhun0_Tg0WoLFMpxThcPbvGgyW6kBaozczNzGv98eRSbJRoTb6WFhvID3kd5vwdwalJpteV5bAiRMSuljSU1NtaaEU2oFrImcb0Vg0E-HMq7JThrYmGstf7wmT3HS_8v30z1O7rKLpwoz4VwBvqyELyO1Wr8KZhAQmYiEAulibzodLvuG5wJSJxl6jQXz8O5IHw8R39IqvJjJ_bipbf5v45twUZQI6NOjfs2LNnJDqwvkAvuwrBfPE9nIcwy7o8mo5dQiDrjR_dk_vQSOZU1uhrh1oFHYnDriwau9sA-1oU-Mvh_Rj2flSc034OH3vV99yYOeRRinTE2j7OCJaYSuU7RPHNIpJXwuhfPDGelMG50khI9_VlJREVpYVG28dJQUrGioPuwMplO7AFErEy4ppW2xNltFZdF6QnoNNekSpk1LSDfA6x0IBnHXBdj5Y2NRKoaFYWoqIBKC86aRq81x8bf1S8RuaYqEmT7Gw4SFdabsphK29DUTVDCcikKJiqdpDoxSVkwTluwizA2LwkItqD9PQ9UWMxvijDB8gyJ9A5_b3UEa9jB2jPThpX57N0ew6r-mI_eZid-nn4Bf6rmTw priority: 102 providerName: IEEE
Title	Majorization-Minimization Algorithm for Discriminative Non-Negative Matrix Factorization
URI	https://ieeexplore.ieee.org/document/9298770 https://www.proquest.com/docview/2474859434 https://doaj.org/article/e0440d31cca24897a47fc01c0d0ba463
Volume	8
WOSCitedRecordID	wos000604520500001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVAON databaseName: DOAJ Directory of Open Access Journals customDbUrl: eissn: 2169-3536 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0000816957 issn: 2169-3536 databaseCode: DOA dateStart: 20130101 isFulltext: true titleUrlDefault: https://www.doaj.org/ providerName: Directory of Open Access Journals – providerCode: PRVHPJ databaseName: ROAD: Directory of Open Access Scholarly Resources customDbUrl: eissn: 2169-3536 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0000816957 issn: 2169-3536 databaseCode: M~E dateStart: 20130101 isFulltext: true titleUrlDefault: https://road.issn.org providerName: ISSN International Centre
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3PT8MgFCbGeNCDUaexOpcePNqMUgrlOOcWD7bxoMlupAU6a7bObNN48m8XKFuamOjFSxNaKOXj9f0g8D0ArmXIlFAkCSRCLMAFUwGLpAqEwEigSFDWkLg-0CxLJhP22Er1ZfaENfTADXB9ZXIiyyjUPSGcMJpjWgoYCihhkWNieT4hZa1gyurgJCQspo5mKISsPxgO9Yh0QIh0nKr9GMvK2TJFlrHfpVj5oZetsRkfgUPnJfqD5uuOwY6qT8BBizuwAyZp_rpYulOUQVrV1dwV_MFsqp-sX-a-9kj9u8poBrPjxWg2P9O1MzVtCqkh6P_0xzbpjmt-Cp7Ho6fhfeDSJAQixngdxLkGqKSJCE30pYEOS2pdKxJLggsq9XBhYRby4wLRMopyZUwXKWSESpzn0RnYrRe1Ogc-LiARUSkU0mFZSVheWH45QQQqQ6ykB9AGMS4ch7hJZTHjNpaAjDcwcwMzdzB74Gbb6K2h0Pi9-q2Zim1Vw39tb2ip4E4q-F9S4YGOmcjtS7QTmFAKPdDdTCx3_-qKI0xxEhuevIv_6PoS7JvhNMs0XbC7Xr6rK7AnPtbVatmzYqqv6deoZw8bfgMr8usx
linkProvider	Directory of Open Access Journals
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PT9swFH6qYBLsMAYFrYONHDgS6thOHB9Lt4ppbcQBpN6sxHagCNqplGl__vwcN0ICIXGLEzty3mf7_Yj9PYATk0irbZbHhlIZ80raWDJjY6051ZRpIRsS17Eoinw6lZcdOG3Pwlhr_eYze4aX_l--WegnDJX1nSrPhXAO-mbKOSXNaa02ooIpJGQqArVQQmR_MBy6r3BOIHW-qbNdPBPnM_XjWfpDWpUXa7FXMKOd93XtM3wKhmQ0aJDfhY6d78HHZ_SCXZhOyrvFMhy0jCez-ewhFKLB_Y17srp9iJzRGv2Y4eKBm2Jw8YsKV7uwN01hghz-_6KRz8sTmu_D9ejn1fAiDpkUYu0ktYrTkhNTi1wn6KA5LJJaeOsrS03GK2GcdEiFsf60oqJmrLSo3bLKMFrzsmQHsDFfzO0XiHhFMs1qbanz3OpMlpWnoNOZpnXCrekBXQtY6UAzjtku7pV3N4hUDSoKUVEBlR6cto3-NCwbb1c_R-TaqkiR7W84SFSYccpiMm3DEjdEKc-lKLmoNUk0MaQqecZ60EUY25cEBHtwtB4HKkznR0W54HmKVHpfX291DFsXV5OxGv8qfh_CNna2idMcwcZq-WS_wQf9dzV7XH73Y_Y_nkTplg
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Majorization-Minimization+Algorithm+for+Discriminative+Non-Negative+Matrix+Factorization&rft.jtitle=IEEE+access&rft.au=Li%2C+Li&rft.au=Kameoka%2C+Hirokazu&rft.au=Makino%2C+Shoji&rft.date=2020&rft.issn=2169-3536&rft.eissn=2169-3536&rft.volume=8&rft.spage=227399&rft.epage=227408&rft_id=info:doi/10.1109%2FACCESS.2020.3045791&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_ACCESS_2020_3045791
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2169-3536&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2169-3536&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2169-3536&client=summon