University of Hertfordshire

By the same authors

Effectiveness in Open-Set Speaker Identification

Research output: Contribution to conferencePaperpeer-review

Standard

Effectiveness in Open-Set Speaker Identification. / Karadaghi, Rawande; Hertlein, Heinz; Ariyaeeinia, A.

2014. Paper presented at 48th IEEE International Carnahan Conference on Security Technology, Rome, Italy.

Research output: Contribution to conferencePaperpeer-review

Harvard

Karadaghi, R, Hertlein, H & Ariyaeeinia, A 2014, 'Effectiveness in Open-Set Speaker Identification', Paper presented at 48th IEEE International Carnahan Conference on Security Technology, Rome, Italy, 13/10/14 - 16/10/14. https://doi.org/10.1109/CCST.2014.6986991

APA

Karadaghi, R., Hertlein, H., & Ariyaeeinia, A. (2014). Effectiveness in Open-Set Speaker Identification. Paper presented at 48th IEEE International Carnahan Conference on Security Technology, Rome, Italy. https://doi.org/10.1109/CCST.2014.6986991

Vancouver

Karadaghi R, Hertlein H, Ariyaeeinia A. Effectiveness in Open-Set Speaker Identification. 2014. Paper presented at 48th IEEE International Carnahan Conference on Security Technology, Rome, Italy. https://doi.org/10.1109/CCST.2014.6986991

Author

Karadaghi, Rawande ; Hertlein, Heinz ; Ariyaeeinia, A. / Effectiveness in Open-Set Speaker Identification. Paper presented at 48th IEEE International Carnahan Conference on Security Technology, Rome, Italy.6 p.

Bibtex

@conference{a6d815c4712e4449b2d3b90d8882253e,
title = "Effectiveness in Open-Set Speaker Identification",
abstract = "This paper presents investigations into the relative effectiveness of two alternative approaches to open-set text-independent speaker identification (OSTI-SI). The methods considered are the recently introduced i-vector and the more traditional GMM-UBM method supported by score normalisation. The study is motivated by the growing need for effective extraction of intelligence and evidence from audio recordings in the fight against crime. OSTI-SI is known to be the most challenging subclass of speaker recognition, and its adoption in criminal investigation applications is further complicated by undesired variations in speech characteristics due to changing levels of environmental noise. In this study, the experimental investigations are conducted using a protocol developed for the identification task, based on the NIST speaker recognition evaluation corpus of 2008. In order to closely cover relevant conditions in the considered application areas and investigate the identification performance in such scenarios, the speech data is contaminated with a range of real-world noise. The paper provides a detailed description of the experimental study and presents a thorough analysis of the results.",
keywords = "Open-set speaker identification, GMM-UBM, i-vector",
author = "Rawande Karadaghi and Heinz Hertlein and A. Ariyaeeinia",
year = "2014",
month = oct,
doi = "10.1109/CCST.2014.6986991",
language = "English",
note = "48th IEEE International Carnahan Conference on Security Technology ; Conference date: 13-10-2014 Through 16-10-2014",

}

RIS

TY - CONF

T1 - Effectiveness in Open-Set Speaker Identification

AU - Karadaghi, Rawande

AU - Hertlein, Heinz

AU - Ariyaeeinia, A.

PY - 2014/10

Y1 - 2014/10

N2 - This paper presents investigations into the relative effectiveness of two alternative approaches to open-set text-independent speaker identification (OSTI-SI). The methods considered are the recently introduced i-vector and the more traditional GMM-UBM method supported by score normalisation. The study is motivated by the growing need for effective extraction of intelligence and evidence from audio recordings in the fight against crime. OSTI-SI is known to be the most challenging subclass of speaker recognition, and its adoption in criminal investigation applications is further complicated by undesired variations in speech characteristics due to changing levels of environmental noise. In this study, the experimental investigations are conducted using a protocol developed for the identification task, based on the NIST speaker recognition evaluation corpus of 2008. In order to closely cover relevant conditions in the considered application areas and investigate the identification performance in such scenarios, the speech data is contaminated with a range of real-world noise. The paper provides a detailed description of the experimental study and presents a thorough analysis of the results.

AB - This paper presents investigations into the relative effectiveness of two alternative approaches to open-set text-independent speaker identification (OSTI-SI). The methods considered are the recently introduced i-vector and the more traditional GMM-UBM method supported by score normalisation. The study is motivated by the growing need for effective extraction of intelligence and evidence from audio recordings in the fight against crime. OSTI-SI is known to be the most challenging subclass of speaker recognition, and its adoption in criminal investigation applications is further complicated by undesired variations in speech characteristics due to changing levels of environmental noise. In this study, the experimental investigations are conducted using a protocol developed for the identification task, based on the NIST speaker recognition evaluation corpus of 2008. In order to closely cover relevant conditions in the considered application areas and investigate the identification performance in such scenarios, the speech data is contaminated with a range of real-world noise. The paper provides a detailed description of the experimental study and presents a thorough analysis of the results.

KW - Open-set speaker identification

KW - GMM-UBM

KW - i-vector

U2 - 10.1109/CCST.2014.6986991

DO - 10.1109/CCST.2014.6986991

M3 - Paper

T2 - 48th IEEE International Carnahan Conference on Security Technology

Y2 - 13 October 2014 through 16 October 2014

ER -