Speaker verification using speaker-specific prompts

Yongxin Zhang, Adel Iskander Fahmy, Michael S Scordilis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Intra- and inter-speaker information, which include acoustical, speaker style, speech rate and temporal variation, despite their critical importance for the verification of claims, still have not been captured effectively. As a result of such modeling deficiency, existing speaker verification systems generally test claimed utterances with interfacing procedures that are common to all speakers. In this paper, a novel method is introduced in which speaker-specific attributes are expressed with reliable, first and second order intra-speaker and inter-speaker statistical information on the output space of speaker models in an explicit way. This is achieved through the computation of the Speech Unit Confusion Matrix (SUCM) that is employed in the scoring phase. An online updating procedure of SUCM is also presented. Experimental results with spoken alphabetic characters used as the basic speech unit indicate that the new method can improve system performance significantly. The method can also be directly extended to the use of other speech units (phonemes, sub-words, digits).

Original languageEnglish
Title of host publicationProceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004
EditorsV. Barr, Z. Markov
Pages826-830
Number of pages5
Volume2
StatePublished - Dec 17 2004
EventProceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004 - Miami Beach, FL, United States
Duration: May 17 2004May 19 2004

Other

OtherProceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004
CountryUnited States
CityMiami Beach, FL
Period5/17/045/19/04

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Zhang, Y., Fahmy, A. I., & Scordilis, M. S. (2004). Speaker verification using speaker-specific prompts. In V. Barr, & Z. Markov (Eds.), Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004 (Vol. 2, pp. 826-830)

Speaker verification using speaker-specific prompts. / Zhang, Yongxin; Fahmy, Adel Iskander; Scordilis, Michael S.

Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004. ed. / V. Barr; Z. Markov. Vol. 2 2004. p. 826-830.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhang, Y, Fahmy, AI & Scordilis, MS 2004, Speaker verification using speaker-specific prompts. in V Barr & Z Markov (eds), Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004. vol. 2, pp. 826-830, Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004, Miami Beach, FL, United States, 5/17/04.
Zhang Y, Fahmy AI, Scordilis MS. Speaker verification using speaker-specific prompts. In Barr V, Markov Z, editors, Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004. Vol. 2. 2004. p. 826-830
Zhang, Yongxin ; Fahmy, Adel Iskander ; Scordilis, Michael S. / Speaker verification using speaker-specific prompts. Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004. editor / V. Barr ; Z. Markov. Vol. 2 2004. pp. 826-830
@inproceedings{49d9074b331e44fcbac93f2e0f4914b5,
title = "Speaker verification using speaker-specific prompts",
abstract = "Intra- and inter-speaker information, which include acoustical, speaker style, speech rate and temporal variation, despite their critical importance for the verification of claims, still have not been captured effectively. As a result of such modeling deficiency, existing speaker verification systems generally test claimed utterances with interfacing procedures that are common to all speakers. In this paper, a novel method is introduced in which speaker-specific attributes are expressed with reliable, first and second order intra-speaker and inter-speaker statistical information on the output space of speaker models in an explicit way. This is achieved through the computation of the Speech Unit Confusion Matrix (SUCM) that is employed in the scoring phase. An online updating procedure of SUCM is also presented. Experimental results with spoken alphabetic characters used as the basic speech unit indicate that the new method can improve system performance significantly. The method can also be directly extended to the use of other speech units (phonemes, sub-words, digits).",
author = "Yongxin Zhang and Fahmy, {Adel Iskander} and Scordilis, {Michael S}",
year = "2004",
month = "12",
day = "17",
language = "English",
isbn = "1577352017",
volume = "2",
pages = "826--830",
editor = "V. Barr and Z. Markov",
booktitle = "Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004",

}

TY - GEN

T1 - Speaker verification using speaker-specific prompts

AU - Zhang, Yongxin

AU - Fahmy, Adel Iskander

AU - Scordilis, Michael S

PY - 2004/12/17

Y1 - 2004/12/17

N2 - Intra- and inter-speaker information, which include acoustical, speaker style, speech rate and temporal variation, despite their critical importance for the verification of claims, still have not been captured effectively. As a result of such modeling deficiency, existing speaker verification systems generally test claimed utterances with interfacing procedures that are common to all speakers. In this paper, a novel method is introduced in which speaker-specific attributes are expressed with reliable, first and second order intra-speaker and inter-speaker statistical information on the output space of speaker models in an explicit way. This is achieved through the computation of the Speech Unit Confusion Matrix (SUCM) that is employed in the scoring phase. An online updating procedure of SUCM is also presented. Experimental results with spoken alphabetic characters used as the basic speech unit indicate that the new method can improve system performance significantly. The method can also be directly extended to the use of other speech units (phonemes, sub-words, digits).

AB - Intra- and inter-speaker information, which include acoustical, speaker style, speech rate and temporal variation, despite their critical importance for the verification of claims, still have not been captured effectively. As a result of such modeling deficiency, existing speaker verification systems generally test claimed utterances with interfacing procedures that are common to all speakers. In this paper, a novel method is introduced in which speaker-specific attributes are expressed with reliable, first and second order intra-speaker and inter-speaker statistical information on the output space of speaker models in an explicit way. This is achieved through the computation of the Speech Unit Confusion Matrix (SUCM) that is employed in the scoring phase. An online updating procedure of SUCM is also presented. Experimental results with spoken alphabetic characters used as the basic speech unit indicate that the new method can improve system performance significantly. The method can also be directly extended to the use of other speech units (phonemes, sub-words, digits).

UR - http://www.scopus.com/inward/record.url?scp=10044219790&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10044219790&partnerID=8YFLogxK

M3 - Conference contribution

SN - 1577352017

SN - 9781577352013

VL - 2

SP - 826

EP - 830

BT - Proceedings of the Seventeenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2004

A2 - Barr, V.

A2 - Markov, Z.

ER -