Loading…

Advances in confidence measures for large vocabulary

This paper addresses the correct choice and combination of confidence measures in large vocabulary speech recognition tasks. We classify single words within continuous as well as large vocabulary utterances into two categories: utterances within the vocabulary which are recognized correctly, and oth...

Full description

Saved in:
Bibliographic Details
Main Authors: Wendemuth, A., Rose, G., Dolfing, J.G.A.
Format: Conference Proceeding
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 708 vol.2
container_issue
container_start_page 705
container_title
container_volume 2
creator Wendemuth, A.
Rose, G.
Dolfing, J.G.A.
description This paper addresses the correct choice and combination of confidence measures in large vocabulary speech recognition tasks. We classify single words within continuous as well as large vocabulary utterances into two categories: utterances within the vocabulary which are recognized correctly, and other utterances, namely misrecognized utterances or (less frequent) out-of-vocabulary (OOV). To this end, we investigate the classification error rate (CER) of several classes of confidence measures and transformations. In particular, we employed data-independent and data-dependent measures. The transformations we investigated include mapping to single confidence measures and linear combinations of these measures. These combinations are computed by means of neural networks trained with Bayes-optimal, and with Gardner-Derrida-optimal criteria. Compared to a recognition system without confidence measures, the selection of (various combinations of) confidence measures, the selection of suitable neural network architectures and training methods, continuously improves the CER.
doi_str_mv 10.1109/ICASSP.1999.759764
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_759764</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>759764</ieee_id><sourcerecordid>759764</sourcerecordid><originalsourceid>FETCH-LOGICAL-i203t-58d237fb708090d6b709c5bfc415c5c80d70e0d65f23858ca67cf6383da9757b3</originalsourceid><addsrcrecordid>eNotj91KAzEQhYM_4Fr7Ar3aF8g62WSSzGUpaoWCQhW8K9n8SKTdlV1b8O0N1Lk553wXM2cYWwhohAC6f14tt9vXRhBRY5CMVhesaqUhLgg-LtktGAsSQQl5xSqBLXAtFN2w-TR9QRmFCEZWTC3DyfU-TnXuaz_0KYdYYn2IbjqOBadhrPdu_Iz1afCuOxb_e8euk9tPcf6vM_b--PC2WvPNy1NptuG5BfnD0YZSKXUGLBAEXQx57JJXAj16C8FALBxTKy1a77TxSUsrgyODppMztjjvzTHG3feYD-X47vyv_AMRa0cL</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Advances in confidence measures for large vocabulary</title><source>IEEE Xplore All Conference Series</source><creator>Wendemuth, A. ; Rose, G. ; Dolfing, J.G.A.</creator><creatorcontrib>Wendemuth, A. ; Rose, G. ; Dolfing, J.G.A.</creatorcontrib><description>This paper addresses the correct choice and combination of confidence measures in large vocabulary speech recognition tasks. We classify single words within continuous as well as large vocabulary utterances into two categories: utterances within the vocabulary which are recognized correctly, and other utterances, namely misrecognized utterances or (less frequent) out-of-vocabulary (OOV). To this end, we investigate the classification error rate (CER) of several classes of confidence measures and transformations. In particular, we employed data-independent and data-dependent measures. The transformations we investigated include mapping to single confidence measures and linear combinations of these measures. These combinations are computed by means of neural networks trained with Bayes-optimal, and with Gardner-Derrida-optimal criteria. Compared to a recognition system without confidence measures, the selection of (various combinations of) confidence measures, the selection of suitable neural network architectures and training methods, continuously improves the CER.</description><identifier>ISSN: 1520-6149</identifier><identifier>ISBN: 0780350413</identifier><identifier>ISBN: 9780780350410</identifier><identifier>EISSN: 2379-190X</identifier><identifier>DOI: 10.1109/ICASSP.1999.759764</identifier><language>eng</language><publisher>IEEE</publisher><subject>Computer architecture ; Computer networks ; Error analysis ; Hidden Markov models ; Laboratories ; Neural networks ; Particle measurements ; Speech recognition ; Vectors ; Vocabulary</subject><ispartof>1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258), 1999, Vol.2, p.705-708 vol.2</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/759764$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2058,4050,4051,27925,54555,54920,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/759764$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Wendemuth, A.</creatorcontrib><creatorcontrib>Rose, G.</creatorcontrib><creatorcontrib>Dolfing, J.G.A.</creatorcontrib><title>Advances in confidence measures for large vocabulary</title><title>1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258)</title><addtitle>ICASSP</addtitle><description>This paper addresses the correct choice and combination of confidence measures in large vocabulary speech recognition tasks. We classify single words within continuous as well as large vocabulary utterances into two categories: utterances within the vocabulary which are recognized correctly, and other utterances, namely misrecognized utterances or (less frequent) out-of-vocabulary (OOV). To this end, we investigate the classification error rate (CER) of several classes of confidence measures and transformations. In particular, we employed data-independent and data-dependent measures. The transformations we investigated include mapping to single confidence measures and linear combinations of these measures. These combinations are computed by means of neural networks trained with Bayes-optimal, and with Gardner-Derrida-optimal criteria. Compared to a recognition system without confidence measures, the selection of (various combinations of) confidence measures, the selection of suitable neural network architectures and training methods, continuously improves the CER.</description><subject>Computer architecture</subject><subject>Computer networks</subject><subject>Error analysis</subject><subject>Hidden Markov models</subject><subject>Laboratories</subject><subject>Neural networks</subject><subject>Particle measurements</subject><subject>Speech recognition</subject><subject>Vectors</subject><subject>Vocabulary</subject><issn>1520-6149</issn><issn>2379-190X</issn><isbn>0780350413</isbn><isbn>9780780350410</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>1999</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNotj91KAzEQhYM_4Fr7Ar3aF8g62WSSzGUpaoWCQhW8K9n8SKTdlV1b8O0N1Lk553wXM2cYWwhohAC6f14tt9vXRhBRY5CMVhesaqUhLgg-LtktGAsSQQl5xSqBLXAtFN2w-TR9QRmFCEZWTC3DyfU-TnXuaz_0KYdYYn2IbjqOBadhrPdu_Iz1afCuOxb_e8euk9tPcf6vM_b--PC2WvPNy1NptuG5BfnD0YZSKXUGLBAEXQx57JJXAj16C8FALBxTKy1a77TxSUsrgyODppMztjjvzTHG3feYD-X47vyv_AMRa0cL</recordid><startdate>1999</startdate><enddate>1999</enddate><creator>Wendemuth, A.</creator><creator>Rose, G.</creator><creator>Dolfing, J.G.A.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>1999</creationdate><title>Advances in confidence measures for large vocabulary</title><author>Wendemuth, A. ; Rose, G. ; Dolfing, J.G.A.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i203t-58d237fb708090d6b709c5bfc415c5c80d70e0d65f23858ca67cf6383da9757b3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>1999</creationdate><topic>Computer architecture</topic><topic>Computer networks</topic><topic>Error analysis</topic><topic>Hidden Markov models</topic><topic>Laboratories</topic><topic>Neural networks</topic><topic>Particle measurements</topic><topic>Speech recognition</topic><topic>Vectors</topic><topic>Vocabulary</topic><toplevel>online_resources</toplevel><creatorcontrib>Wendemuth, A.</creatorcontrib><creatorcontrib>Rose, G.</creatorcontrib><creatorcontrib>Dolfing, J.G.A.</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 Electronic Library Online</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Wendemuth, A.</au><au>Rose, G.</au><au>Dolfing, J.G.A.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Advances in confidence measures for large vocabulary</atitle><btitle>1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258)</btitle><stitle>ICASSP</stitle><date>1999</date><risdate>1999</risdate><volume>2</volume><spage>705</spage><epage>708 vol.2</epage><pages>705-708 vol.2</pages><issn>1520-6149</issn><eissn>2379-190X</eissn><isbn>0780350413</isbn><isbn>9780780350410</isbn><abstract>This paper addresses the correct choice and combination of confidence measures in large vocabulary speech recognition tasks. We classify single words within continuous as well as large vocabulary utterances into two categories: utterances within the vocabulary which are recognized correctly, and other utterances, namely misrecognized utterances or (less frequent) out-of-vocabulary (OOV). To this end, we investigate the classification error rate (CER) of several classes of confidence measures and transformations. In particular, we employed data-independent and data-dependent measures. The transformations we investigated include mapping to single confidence measures and linear combinations of these measures. These combinations are computed by means of neural networks trained with Bayes-optimal, and with Gardner-Derrida-optimal criteria. Compared to a recognition system without confidence measures, the selection of (various combinations of) confidence measures, the selection of suitable neural network architectures and training methods, continuously improves the CER.</abstract><pub>IEEE</pub><doi>10.1109/ICASSP.1999.759764</doi></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1520-6149
ispartof 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258), 1999, Vol.2, p.705-708 vol.2
issn 1520-6149
2379-190X
language eng
recordid cdi_ieee_primary_759764
source IEEE Xplore All Conference Series
subjects Computer architecture
Computer networks
Error analysis
Hidden Markov models
Laboratories
Neural networks
Particle measurements
Speech recognition
Vectors
Vocabulary
title Advances in confidence measures for large vocabulary
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T12%3A27%3A28IST&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=Advances%20in%20confidence%20measures%20for%20large%20vocabulary&rft.btitle=1999%20IEEE%20International%20Conference%20on%20Acoustics,%20Speech,%20and%20Signal%20Processing.%20Proceedings.%20ICASSP99%20(Cat.%20No.99CH36258)&rft.au=Wendemuth,%20A.&rft.date=1999&rft.volume=2&rft.spage=705&rft.epage=708%20vol.2&rft.pages=705-708%20vol.2&rft.issn=1520-6149&rft.eissn=2379-190X&rft.isbn=0780350413&rft.isbn_list=9780780350410&rft_id=info:doi/10.1109/ICASSP.1999.759764&rft_dat=%3Cieee_CHZPO%3E759764%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i203t-58d237fb708090d6b709c5bfc415c5c80d70e0d65f23858ca67cf6383da9757b3%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=759764&rfr_iscdi=true