Kernel Additive Models for Source Separation
Source separation consists of separating a signal into additive components. It is a topic of considerable interest with many applications that has gathered much attention recently. Here, we introduce a new framework for source separation called Kernel Additive Modelling, which is based on local regr...
Gespeichert in:
Veröffentlicht in: | IEEE transactions on signal processing 2014-08, Vol.62 (16), p.4298-4310 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 4310 |
---|---|
container_issue | 16 |
container_start_page | 4298 |
container_title | IEEE transactions on signal processing |
container_volume | 62 |
creator | Liutkus, Antoine Fitzgerald, Derry Rafii, Zafar Pardo, Bryan Daudet, Laurent |
description | Source separation consists of separating a signal into additive components. It is a topic of considerable interest with many applications that has gathered much attention recently. Here, we introduce a new framework for source separation called Kernel Additive Modelling, which is based on local regression and permits efficient separation of multidimensional and/or nonnegative and/or non-regularly sampled signals. The main idea of the method is to assume that a source at some location can be estimated using its values at other locations nearby, where nearness is defined through a source-specific proximity kernel. Such a kernel provides an efficient way to account for features like periodicity, continuity, smoothness, stability over time or frequency, and self-similarity. In many cases, such local dynamics are indeed much more natural to assess than any global model such as a tensor factorization. This framework permits one to use different proximity kernels for different sources and to separate them using the iterative kernel backfitting algorithm we describe. As we show, kernel additive modelling generalizes many recent and efficient techniques for source separation and opens the path to creating and combining source models in a principled way. Experimental results on the separation of synthetic and audio signals demonstrate the effectiveness of the approach. |
doi_str_mv | 10.1109/TSP.2014.2332434 |
format | Article |
fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_proquest_miscellaneous_1559709677</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>6842708</ieee_id><sourcerecordid>3443565381</sourcerecordid><originalsourceid>FETCH-LOGICAL-c430t-62e4e0f57f6b575a799a784b94d44041fe42ea89d491c86362b6403e1ff4144a3</originalsourceid><addsrcrecordid>eNpd0E1LAzEQBuBFFKzVu-BlQQQFt-Zj8nUsRa1YUWgFbyHdneDKuluTtuC_N6WlB08JyTPDzJtl55QMKCXmbjZ9GzBCYcA4Z8DhIOtRA7QgoORhuhPBC6HVx3F2EuMXSRKM7GW3zxhabPJhVdXLeo35S1dhE3PfhXzarUKJ-RQXLrhl3bWn2ZF3TcSz3dnP3h_uZ6NxMXl9fBoNJ0UJnCwLyRCQeKG8nAslnDLGKQ1zAxUAAeoRGDptKjC01JJLNpdAOFLvIY3leD-72fb9dI1dhPrbhV_budqOhxO7eSOUpKUB1izZ661dhO5nhXFpv-tYYtO4FrtVtFQIo4iRSiV6-Y9-pQ3btElSUgjOpdZJka0qQxdjQL-fgBK7idqmqO0maruLOpVc7Rq7WLrGB9eWddzXMa0Y1dQkd7F1NSLuv6UGpojmfwGggtE</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>1565533688</pqid></control><display><type>article</type><title>Kernel Additive Models for Source Separation</title><source>IEEE Xplore</source><creator>Liutkus, Antoine ; Fitzgerald, Derry ; Rafii, Zafar ; Pardo, Bryan ; Daudet, Laurent</creator><creatorcontrib>Liutkus, Antoine ; Fitzgerald, Derry ; Rafii, Zafar ; Pardo, Bryan ; Daudet, Laurent</creatorcontrib><description>Source separation consists of separating a signal into additive components. It is a topic of considerable interest with many applications that has gathered much attention recently. Here, we introduce a new framework for source separation called Kernel Additive Modelling, which is based on local regression and permits efficient separation of multidimensional and/or nonnegative and/or non-regularly sampled signals. The main idea of the method is to assume that a source at some location can be estimated using its values at other locations nearby, where nearness is defined through a source-specific proximity kernel. Such a kernel provides an efficient way to account for features like periodicity, continuity, smoothness, stability over time or frequency, and self-similarity. In many cases, such local dynamics are indeed much more natural to assess than any global model such as a tensor factorization. This framework permits one to use different proximity kernels for different sources and to separate them using the iterative kernel backfitting algorithm we describe. As we show, kernel additive modelling generalizes many recent and efficient techniques for source separation and opens the path to creating and combining source models in a principled way. Experimental results on the separation of synthetic and audio signals demonstrate the effectiveness of the approach.</description><identifier>ISSN: 1053-587X</identifier><identifier>EISSN: 1941-0476</identifier><identifier>DOI: 10.1109/TSP.2014.2332434</identifier><identifier>CODEN: ITPRED</identifier><language>eng</language><publisher>New York, NY: IEEE</publisher><subject>Adaptation models ; Additives ; Applied sciences ; Audio signals ; Composting ; Computer Science ; Context ; Detection, estimation, filtering, equalization, prediction ; Engineering Sciences ; Exact sciences and technology ; Heuristic algorithms ; Information, signal and communications theory ; Kernel ; kernel method ; Kernels ; local regression ; Mathematical models ; Modelling ; nonparametric models ; Parametric statistics ; Proximity ; Separation ; Signal and communications theory ; Signal and Image processing ; Signal, noise ; Source separation ; Telecommunications and information theory</subject><ispartof>IEEE transactions on signal processing, 2014-08, Vol.62 (16), p.4298-4310</ispartof><rights>2015 INIST-CNRS</rights><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2014</rights><rights>Distributed under a Creative Commons Attribution 4.0 International License</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c430t-62e4e0f57f6b575a799a784b94d44041fe42ea89d491c86362b6403e1ff4144a3</citedby><cites>FETCH-LOGICAL-c430t-62e4e0f57f6b575a799a784b94d44041fe42ea89d491c86362b6403e1ff4144a3</cites><orcidid>0000-0002-3458-6498</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/6842708$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>230,314,780,784,796,885,27924,27925,54758</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/6842708$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc><backlink>$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=28721819$$DView record in Pascal Francis$$Hfree_for_read</backlink><backlink>$$Uhttps://inria.hal.science/hal-01011044$$DView record in HAL$$Hfree_for_read</backlink></links><search><creatorcontrib>Liutkus, Antoine</creatorcontrib><creatorcontrib>Fitzgerald, Derry</creatorcontrib><creatorcontrib>Rafii, Zafar</creatorcontrib><creatorcontrib>Pardo, Bryan</creatorcontrib><creatorcontrib>Daudet, Laurent</creatorcontrib><title>Kernel Additive Models for Source Separation</title><title>IEEE transactions on signal processing</title><addtitle>TSP</addtitle><description>Source separation consists of separating a signal into additive components. It is a topic of considerable interest with many applications that has gathered much attention recently. Here, we introduce a new framework for source separation called Kernel Additive Modelling, which is based on local regression and permits efficient separation of multidimensional and/or nonnegative and/or non-regularly sampled signals. The main idea of the method is to assume that a source at some location can be estimated using its values at other locations nearby, where nearness is defined through a source-specific proximity kernel. Such a kernel provides an efficient way to account for features like periodicity, continuity, smoothness, stability over time or frequency, and self-similarity. In many cases, such local dynamics are indeed much more natural to assess than any global model such as a tensor factorization. This framework permits one to use different proximity kernels for different sources and to separate them using the iterative kernel backfitting algorithm we describe. As we show, kernel additive modelling generalizes many recent and efficient techniques for source separation and opens the path to creating and combining source models in a principled way. Experimental results on the separation of synthetic and audio signals demonstrate the effectiveness of the approach.</description><subject>Adaptation models</subject><subject>Additives</subject><subject>Applied sciences</subject><subject>Audio signals</subject><subject>Composting</subject><subject>Computer Science</subject><subject>Context</subject><subject>Detection, estimation, filtering, equalization, prediction</subject><subject>Engineering Sciences</subject><subject>Exact sciences and technology</subject><subject>Heuristic algorithms</subject><subject>Information, signal and communications theory</subject><subject>Kernel</subject><subject>kernel method</subject><subject>Kernels</subject><subject>local regression</subject><subject>Mathematical models</subject><subject>Modelling</subject><subject>nonparametric models</subject><subject>Parametric statistics</subject><subject>Proximity</subject><subject>Separation</subject><subject>Signal and communications theory</subject><subject>Signal and Image processing</subject><subject>Signal, noise</subject><subject>Source separation</subject><subject>Telecommunications and information theory</subject><issn>1053-587X</issn><issn>1941-0476</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2014</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNpd0E1LAzEQBuBFFKzVu-BlQQQFt-Zj8nUsRa1YUWgFbyHdneDKuluTtuC_N6WlB08JyTPDzJtl55QMKCXmbjZ9GzBCYcA4Z8DhIOtRA7QgoORhuhPBC6HVx3F2EuMXSRKM7GW3zxhabPJhVdXLeo35S1dhE3PfhXzarUKJ-RQXLrhl3bWn2ZF3TcSz3dnP3h_uZ6NxMXl9fBoNJ0UJnCwLyRCQeKG8nAslnDLGKQ1zAxUAAeoRGDptKjC01JJLNpdAOFLvIY3leD-72fb9dI1dhPrbhV_budqOhxO7eSOUpKUB1izZ661dhO5nhXFpv-tYYtO4FrtVtFQIo4iRSiV6-Y9-pQ3btElSUgjOpdZJka0qQxdjQL-fgBK7idqmqO0maruLOpVc7Rq7WLrGB9eWddzXMa0Y1dQkd7F1NSLuv6UGpojmfwGggtE</recordid><startdate>20140815</startdate><enddate>20140815</enddate><creator>Liutkus, Antoine</creator><creator>Fitzgerald, Derry</creator><creator>Rafii, Zafar</creator><creator>Pardo, Bryan</creator><creator>Daudet, Laurent</creator><general>IEEE</general><general>Institute of Electrical and Electronics Engineers</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>IQODW</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>F28</scope><scope>FR3</scope><scope>1XC</scope><scope>VOOES</scope><orcidid>https://orcid.org/0000-0002-3458-6498</orcidid></search><sort><creationdate>20140815</creationdate><title>Kernel Additive Models for Source Separation</title><author>Liutkus, Antoine ; Fitzgerald, Derry ; Rafii, Zafar ; Pardo, Bryan ; Daudet, Laurent</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c430t-62e4e0f57f6b575a799a784b94d44041fe42ea89d491c86362b6403e1ff4144a3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2014</creationdate><topic>Adaptation models</topic><topic>Additives</topic><topic>Applied sciences</topic><topic>Audio signals</topic><topic>Composting</topic><topic>Computer Science</topic><topic>Context</topic><topic>Detection, estimation, filtering, equalization, prediction</topic><topic>Engineering Sciences</topic><topic>Exact sciences and technology</topic><topic>Heuristic algorithms</topic><topic>Information, signal and communications theory</topic><topic>Kernel</topic><topic>kernel method</topic><topic>Kernels</topic><topic>local regression</topic><topic>Mathematical models</topic><topic>Modelling</topic><topic>nonparametric models</topic><topic>Parametric statistics</topic><topic>Proximity</topic><topic>Separation</topic><topic>Signal and communications theory</topic><topic>Signal and Image processing</topic><topic>Signal, noise</topic><topic>Source separation</topic><topic>Telecommunications and information theory</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Liutkus, Antoine</creatorcontrib><creatorcontrib>Fitzgerald, Derry</creatorcontrib><creatorcontrib>Rafii, Zafar</creatorcontrib><creatorcontrib>Pardo, Bryan</creatorcontrib><creatorcontrib>Daudet, Laurent</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Xplore</collection><collection>Pascal-Francis</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics & Communications Abstracts</collection><collection>Technology 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>ANTE: Abstracts in New Technology & Engineering</collection><collection>Engineering Research Database</collection><collection>Hyper Article en Ligne (HAL)</collection><collection>Hyper Article en Ligne (HAL) (Open Access)</collection><jtitle>IEEE transactions on signal processing</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Liutkus, Antoine</au><au>Fitzgerald, Derry</au><au>Rafii, Zafar</au><au>Pardo, Bryan</au><au>Daudet, Laurent</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Kernel Additive Models for Source Separation</atitle><jtitle>IEEE transactions on signal processing</jtitle><stitle>TSP</stitle><date>2014-08-15</date><risdate>2014</risdate><volume>62</volume><issue>16</issue><spage>4298</spage><epage>4310</epage><pages>4298-4310</pages><issn>1053-587X</issn><eissn>1941-0476</eissn><coden>ITPRED</coden><abstract>Source separation consists of separating a signal into additive components. It is a topic of considerable interest with many applications that has gathered much attention recently. Here, we introduce a new framework for source separation called Kernel Additive Modelling, which is based on local regression and permits efficient separation of multidimensional and/or nonnegative and/or non-regularly sampled signals. The main idea of the method is to assume that a source at some location can be estimated using its values at other locations nearby, where nearness is defined through a source-specific proximity kernel. Such a kernel provides an efficient way to account for features like periodicity, continuity, smoothness, stability over time or frequency, and self-similarity. In many cases, such local dynamics are indeed much more natural to assess than any global model such as a tensor factorization. This framework permits one to use different proximity kernels for different sources and to separate them using the iterative kernel backfitting algorithm we describe. As we show, kernel additive modelling generalizes many recent and efficient techniques for source separation and opens the path to creating and combining source models in a principled way. Experimental results on the separation of synthetic and audio signals demonstrate the effectiveness of the approach.</abstract><cop>New York, NY</cop><pub>IEEE</pub><doi>10.1109/TSP.2014.2332434</doi><tpages>13</tpages><orcidid>https://orcid.org/0000-0002-3458-6498</orcidid><oa>free_for_read</oa></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | ISSN: 1053-587X |
ispartof | IEEE transactions on signal processing, 2014-08, Vol.62 (16), p.4298-4310 |
issn | 1053-587X 1941-0476 |
language | eng |
recordid | cdi_proquest_miscellaneous_1559709677 |
source | IEEE Xplore |
subjects | Adaptation models Additives Applied sciences Audio signals Composting Computer Science Context Detection, estimation, filtering, equalization, prediction Engineering Sciences Exact sciences and technology Heuristic algorithms Information, signal and communications theory Kernel kernel method Kernels local regression Mathematical models Modelling nonparametric models Parametric statistics Proximity Separation Signal and communications theory Signal and Image processing Signal, noise Source separation Telecommunications and information theory |
title | Kernel Additive Models for Source Separation |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T13%3A04%3A33IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Kernel%20Additive%20Models%20for%20Source%20Separation&rft.jtitle=IEEE%20transactions%20on%20signal%20processing&rft.au=Liutkus,%20Antoine&rft.date=2014-08-15&rft.volume=62&rft.issue=16&rft.spage=4298&rft.epage=4310&rft.pages=4298-4310&rft.issn=1053-587X&rft.eissn=1941-0476&rft.coden=ITPRED&rft_id=info:doi/10.1109/TSP.2014.2332434&rft_dat=%3Cproquest_RIE%3E3443565381%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=1565533688&rft_id=info:pmid/&rft_ieee_id=6842708&rfr_iscdi=true |