Publication detail

Speaker Discrimination Using Long-Term Spectrum of Speech

SIGMUND, M.

Original Title

Speaker Discrimination Using Long-Term Spectrum of Speech

Type

journal article in Web of Science

Language

English

Original Abstract

In this article, a specific long-term speech spectrum was investigated with respect to its use for speaker recognition. The long-term spectrum was calculated by means of second-order linear prediction using the average autocorrelation coefficients. Four subbands with the most discriminative capability were selected for speaker recognition. These subbands involve the frequencies of 0-1.2 kHz in total. The best recognition rates, i.e. 91.7% on complete speech and 100% on voiced speech, were achieved in optimal paired subbands.

Keywords

Speech signal, long-term spectrum, speaker discrimination, efficient features

Authors

SIGMUND, M.

Released

25. 9. 2019

Publisher

Kaunas University of Technology

Location

Kaunas

ISBN

1392-124X

Periodical

Information Technology and Control

Year of study

48

Number

3

State

Republic of Lithuania

Pages from

446

Pages to

453

Pages count

8

URL

http://itc.ktu.lt/index.php/ITC/article/view/21248

Full text in the Digital Library

http://hdl.handle.net/11012/201688

BibTex

@article{BUT159590,
  author="Milan {Sigmund}",
  title="Speaker Discrimination Using Long-Term Spectrum of Speech",
  journal="Information Technology and Control",
  year="2019",
  volume="48",
  number="3",
  pages="446--453",
  doi="10.5755/j01.itc.48.3.21248",
  issn="1392-124X",
  url="http://itc.ktu.lt/index.php/ITC/article/view/21248"
}