Selective use of the speech spectrum and a VQGMM method for speaker identification

Qiguang Lin; Ea Ee Jan; Chi Wei Che; Dong Suk Yuk; James Flanagan

Selective use of the speech spectrum and a VQGMM method for speaker identification

Qiguang Lin, Ea Ee Jan, Chi Wei Che, Dong Suk Yuk, James Flanagan

Research output: Contribution to conference › Paper › peer-review

Abstract

This paper describes two separate sets of speaker identification experiments. In the first set of experiments, the speech spectrum is selectively used for speaker identification. The results show that the higher portion of the speech spectrum contains more reliable idiosyncratic information on speakers than does the lower portion of equal bandwidth. In the second set of experiments, a vector-quantization based Gaussian mixture models (VQGMMs) is developed for text-independent speaker identification. The system has been evaluated in the recent speaker identification evaluation organized by NIST. In this paper, details of the system design are given and the evaluation results are presented.

Original language	English
Pages	2415-2418
Number of pages	4
Publication status	Published - 1996
Externally published	Yes
Event	Proceedings of the 1996 International Conference on Spoken Language Processing, ICSLP. Part 1 (of 4) - Philadelphia, PA, USA Duration: 1996 Oct 3 → 1996 Oct 6

Conference

Conference	Proceedings of the 1996 International Conference on Spoken Language Processing, ICSLP. Part 1 (of 4)
City	Philadelphia, PA, USA
Period	96/10/3 → 96/10/6

ASJC Scopus subject areas

General Computer Science

Cite this

@conference{c20a7f6d88fc4932a458c17b2660f573,

title = "Selective use of the speech spectrum and a VQGMM method for speaker identification",

abstract = "This paper describes two separate sets of speaker identification experiments. In the first set of experiments, the speech spectrum is selectively used for speaker identification. The results show that the higher portion of the speech spectrum contains more reliable idiosyncratic information on speakers than does the lower portion of equal bandwidth. In the second set of experiments, a vector-quantization based Gaussian mixture models (VQGMMs) is developed for text-independent speaker identification. The system has been evaluated in the recent speaker identification evaluation organized by NIST. In this paper, details of the system design are given and the evaluation results are presented.",

author = "Qiguang Lin and Jan, {Ea Ee} and Che, {Chi Wei} and Yuk, {Dong Suk} and James Flanagan",

year = "1996",

language = "English",

pages = "2415--2418",

note = "Proceedings of the 1996 International Conference on Spoken Language Processing, ICSLP. Part 1 (of 4) ; Conference date: 03-10-1996 Through 06-10-1996",

}

TY - CONF

T1 - Selective use of the speech spectrum and a VQGMM method for speaker identification

AU - Lin, Qiguang

AU - Jan, Ea Ee

AU - Che, Chi Wei

AU - Yuk, Dong Suk

AU - Flanagan, James

PY - 1996

Y1 - 1996

N2 - This paper describes two separate sets of speaker identification experiments. In the first set of experiments, the speech spectrum is selectively used for speaker identification. The results show that the higher portion of the speech spectrum contains more reliable idiosyncratic information on speakers than does the lower portion of equal bandwidth. In the second set of experiments, a vector-quantization based Gaussian mixture models (VQGMMs) is developed for text-independent speaker identification. The system has been evaluated in the recent speaker identification evaluation organized by NIST. In this paper, details of the system design are given and the evaluation results are presented.

AB - This paper describes two separate sets of speaker identification experiments. In the first set of experiments, the speech spectrum is selectively used for speaker identification. The results show that the higher portion of the speech spectrum contains more reliable idiosyncratic information on speakers than does the lower portion of equal bandwidth. In the second set of experiments, a vector-quantization based Gaussian mixture models (VQGMMs) is developed for text-independent speaker identification. The system has been evaluated in the recent speaker identification evaluation organized by NIST. In this paper, details of the system design are given and the evaluation results are presented.

UR - http://www.scopus.com/inward/record.url?scp=0030351622&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0030351622

SP - 2415

EP - 2418

T2 - Proceedings of the 1996 International Conference on Spoken Language Processing, ICSLP. Part 1 (of 4)

Y2 - 3 October 1996 through 6 October 1996

ER -

Selective use of the speech spectrum and a VQGMM method for speaker identification

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this