Convolutional Neural Network for speaker change detection in telephone speaker diarization system

The aim of this paper is to propose a speaker change detection technique based on Convolutional Neural Network (CNN) and evaluate its contribution to the performance of a speaker diarization system for telephone conversations. For the comparison we used an i-vector based speaker diarization system....

Full description

Saved in:
Bibliographic Details
Main Authors: Hruz, Marek, Zajic, Zbynek
Format: Conference Proceeding
Language:English
Subjects:
Convolution
Convolutional Neural Network
Generalized Likelihood Ratio
Labeling
Measurement
Speaker Change Detection
Speaker Diarization
Spectrogram
Speech
Telephone sets
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 4949
container_issue
container_start_page 4945
container_title
container_volume
creator Hruz, Marek
Zajic, Zbynek
description The aim of this paper is to propose a speaker change detection technique based on Convolutional Neural Network (CNN) and evaluate its contribution to the performance of a speaker diarization system for telephone conversations. For the comparison we used an i-vector based speaker diarization system. The baseline speaker change detection uses Generalized Likelihood Ratio (GLR) metric. Experiments were conducted on the English part of the CallHome corpus. Our proposed CNN speaker change detection outperformed the GLR approach, reducing the Equal Error Rate relatively by 46 %. The final results on speaker diarization system indicate that the use of speaker change detection based on CNN is beneficial with relative improvement of diarization error rate by 28 %.
doi_str_mv 10.1109/ICASSP.2017.7953097
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_7953097</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>7953097</ieee_id><sourcerecordid>7953097</sourcerecordid><originalsourceid>FETCH-LOGICAL-i175t-6430a4b57807ce729ca4ed2f19160a3329b53b9ff9afeca238374c27ccc1810b3</originalsourceid><addsrcrecordid>eNo9kM1Kw0AURkdBsK0-QTfzAqn3ziSZzFKCP4WiQhXclZvJjR2bJmGSKvXp1Vpcnc3h4_AJMUWYIYK9mufXy-XTTAGambGJBmtOxBgTsBAjmvRUjJQ2NkILr-di3PfvAJCZOBsJytvmo613g28bquUD78IBw2cbNrJqg-w7pg0H6dbUvLEseWD3a0vfyIFr7tZtw_9W6Sn4LzoI_b4feHshziqqe748ciJebm-e8_to8Xj3E76IPJpkiNJYA8VFYjIwjo2yjmIuVYUWUyCtlS0SXdiqslSxI6UzbWKnjHMOM4RCT8T0b9cz86oLfkthvzreob8BtMlW9g</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Convolutional Neural Network for speaker change detection in telephone speaker diarization system</title><source>IEEE Xplore All Conference Series</source><creator>Hruz, Marek ; Zajic, Zbynek</creator><creatorcontrib>Hruz, Marek ; Zajic, Zbynek</creatorcontrib><description>The aim of this paper is to propose a speaker change detection technique based on Convolutional Neural Network (CNN) and evaluate its contribution to the performance of a speaker diarization system for telephone conversations. For the comparison we used an i-vector based speaker diarization system. The baseline speaker change detection uses Generalized Likelihood Ratio (GLR) metric. Experiments were conducted on the English part of the CallHome corpus. Our proposed CNN speaker change detection outperformed the GLR approach, reducing the Equal Error Rate relatively by 46 %. The final results on speaker diarization system indicate that the use of speaker change detection based on CNN is beneficial with relative improvement of diarization error rate by 28 %.</description><identifier>EISSN: 2379-190X</identifier><identifier>EISBN: 1509041176</identifier><identifier>EISBN: 9781509041176</identifier><identifier>DOI: 10.1109/ICASSP.2017.7953097</identifier><language>eng</language><publisher>IEEE</publisher><subject>Convolution ; Convolutional Neural Network ; Generalized Likelihood Ratio ; Labeling ; Measurement ; Speaker Change Detection ; Speaker Diarization ; Spectrogram ; Speech ; Telephone sets</subject><ispartof>2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2017, p.4945-4949</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/7953097$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,23930,23931,25140,27925,54555,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/7953097$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Hruz, Marek</creatorcontrib><creatorcontrib>Zajic, Zbynek</creatorcontrib><title>Convolutional Neural Network for speaker change detection in telephone speaker diarization system</title><title>2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)</title><addtitle>ICASSP</addtitle><description>The aim of this paper is to propose a speaker change detection technique based on Convolutional Neural Network (CNN) and evaluate its contribution to the performance of a speaker diarization system for telephone conversations. For the comparison we used an i-vector based speaker diarization system. The baseline speaker change detection uses Generalized Likelihood Ratio (GLR) metric. Experiments were conducted on the English part of the CallHome corpus. Our proposed CNN speaker change detection outperformed the GLR approach, reducing the Equal Error Rate relatively by 46 %. The final results on speaker diarization system indicate that the use of speaker change detection based on CNN is beneficial with relative improvement of diarization error rate by 28 %.</description><subject>Convolution</subject><subject>Convolutional Neural Network</subject><subject>Generalized Likelihood Ratio</subject><subject>Labeling</subject><subject>Measurement</subject><subject>Speaker Change Detection</subject><subject>Speaker Diarization</subject><subject>Spectrogram</subject><subject>Speech</subject><subject>Telephone sets</subject><issn>2379-190X</issn><isbn>1509041176</isbn><isbn>9781509041176</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2017</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNo9kM1Kw0AURkdBsK0-QTfzAqn3ziSZzFKCP4WiQhXclZvJjR2bJmGSKvXp1Vpcnc3h4_AJMUWYIYK9mufXy-XTTAGambGJBmtOxBgTsBAjmvRUjJQ2NkILr-di3PfvAJCZOBsJytvmo613g28bquUD78IBw2cbNrJqg-w7pg0H6dbUvLEseWD3a0vfyIFr7tZtw_9W6Sn4LzoI_b4feHshziqqe748ciJebm-e8_to8Xj3E76IPJpkiNJYA8VFYjIwjo2yjmIuVYUWUyCtlS0SXdiqslSxI6UzbWKnjHMOM4RCT8T0b9cz86oLfkthvzreob8BtMlW9g</recordid><startdate>201703</startdate><enddate>201703</enddate><creator>Hruz, Marek</creator><creator>Zajic, Zbynek</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>201703</creationdate><title>Convolutional Neural Network for speaker change detection in telephone speaker diarization system</title><author>Hruz, Marek ; Zajic, Zbynek</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i175t-6430a4b57807ce729ca4ed2f19160a3329b53b9ff9afeca238374c27ccc1810b3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2017</creationdate><topic>Convolution</topic><topic>Convolutional Neural Network</topic><topic>Generalized Likelihood Ratio</topic><topic>Labeling</topic><topic>Measurement</topic><topic>Speaker Change Detection</topic><topic>Speaker Diarization</topic><topic>Spectrogram</topic><topic>Speech</topic><topic>Telephone sets</topic><toplevel>online_resources</toplevel><creatorcontrib>Hruz, Marek</creatorcontrib><creatorcontrib>Zajic, Zbynek</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE/IET Electronic Library</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Hruz, Marek</au><au>Zajic, Zbynek</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Convolutional Neural Network for speaker change detection in telephone speaker diarization system</atitle><btitle>2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)</btitle><stitle>ICASSP</stitle><date>2017-03</date><risdate>2017</risdate><spage>4945</spage><epage>4949</epage><pages>4945-4949</pages><eissn>2379-190X</eissn><eisbn>1509041176</eisbn><eisbn>9781509041176</eisbn><abstract>The aim of this paper is to propose a speaker change detection technique based on Convolutional Neural Network (CNN) and evaluate its contribution to the performance of a speaker diarization system for telephone conversations. For the comparison we used an i-vector based speaker diarization system. The baseline speaker change detection uses Generalized Likelihood Ratio (GLR) metric. Experiments were conducted on the English part of the CallHome corpus. Our proposed CNN speaker change detection outperformed the GLR approach, reducing the Equal Error Rate relatively by 46 %. The final results on speaker diarization system indicate that the use of speaker change detection based on CNN is beneficial with relative improvement of diarization error rate by 28 %.</abstract><pub>IEEE</pub><doi>10.1109/ICASSP.2017.7953097</doi><tpages>5</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier EISSN: 2379-190X
ispartof 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2017, p.4945-4949
issn 2379-190X
language eng
recordid cdi_ieee_primary_7953097
source IEEE Xplore All Conference Series
subjects Convolution
Convolutional Neural Network
Generalized Likelihood Ratio
Labeling
Measurement
Speaker Change Detection
Speaker Diarization
Spectrogram
Speech
Telephone sets
title Convolutional Neural Network for speaker change detection in telephone speaker diarization system
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-04T22%3A18%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_CHZPO&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Convolutional%20Neural%20Network%20for%20speaker%20change%20detection%20in%20telephone%20speaker%20diarization%20system&rft.btitle=2017%20IEEE%20International%20Conference%20on%20Acoustics,%20Speech%20and%20Signal%20Processing%20(ICASSP)&rft.au=Hruz,%20Marek&rft.date=2017-03&rft.spage=4945&rft.epage=4949&rft.pages=4945-4949&rft.eissn=2379-190X&rft_id=info:doi/10.1109/ICASSP.2017.7953097&rft.eisbn=1509041176&rft.eisbn_list=9781509041176&rft_dat=%3Cieee_CHZPO%3E7953097%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i175t-6430a4b57807ce729ca4ed2f19160a3329b53b9ff9afeca238374c27ccc1810b3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=7953097&rfr_iscdi=true