Toward open-set text-independent speaker identification in tactical communications

Matt B. Wolf; Won Kyung Park; Jae C. Oh; Misty K. Blowers

doi:10.1109/CISDA.2007.368129

Toward open-set text-independent speaker identification in tactical communications

Matt B. Wolf, Won Kyung Park, Jae C. Oh, Misty K. Blowers

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Entry/Poem › Conference contribution

2 Scopus citations

Abstract

We present the design and implementation of an open-set textindependent speaker identification system using genetic Learning Classifier Systems (LCS). We examine the use of this system in a real-number problem domain, where there is strong interest in its application to tactical communications. We investigate different encoding methods for representing real-number knowledge and study the efficacy of each method for speaker identification. We also identify several difficulties in solving the speaker identification problems with LCS and introduce new approaches to resolve the difficulties. Experimental results show that our system successfully learns 200 voice features at accuracies of 90 % to 100 % and 15,000 features to more than 80% for the closed-set problem, which is considered a strong result in the speaker identification community. The open-set capability is also comparable to existing numeric-based methods.

Original language	English (US)
Title of host publication	Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007
Pages	7-14
Number of pages	8
DOIs	https://doi.org/10.1109/CISDA.2007.368129
State	Published - 2007
Event	2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007 - Honolulu, HI, United States Duration: Apr 1 2007 → Apr 5 2007

Publication series

Name	Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007

Other

Other	2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007
Country/Territory	United States
City	Honolulu, HI
Period	4/1/07 → 4/5/07

Keywords

Classifier systems
Genetic algorithms
Language and speech
Machine learning

ASJC Scopus subject areas

Artificial Intelligence
Electrical and Electronic Engineering
Computational Mathematics
Theoretical Computer Science

Access to Document

10.1109/CISDA.2007.368129

Cite this

Wolf, M. B., Park, W. K., Oh, J. C., & Blowers, M. K. (2007). Toward open-set text-independent speaker identification in tactical communications. In Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007 (pp. 7-14). Article 4219076 (Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007). https://doi.org/10.1109/CISDA.2007.368129

Toward open-set text-independent speaker identification in tactical communications. / Wolf, Matt B.; Park, Won Kyung; Oh, Jae C. et al.
Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007. 2007. p. 7-14 4219076 (Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Wolf, MB, Park, WK, Oh, JC & Blowers, MK 2007, Toward open-set text-independent speaker identification in tactical communications. in Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007., 4219076, Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007, pp. 7-14, 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007, Honolulu, HI, United States, 4/1/07. https://doi.org/10.1109/CISDA.2007.368129

Wolf MB, Park WK, Oh JC, Blowers MK. Toward open-set text-independent speaker identification in tactical communications. In Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007. 2007. p. 7-14. 4219076. (Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007). doi: 10.1109/CISDA.2007.368129

Wolf, Matt B. ; Park, Won Kyung ; Oh, Jae C. et al. / Toward open-set text-independent speaker identification in tactical communications. Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007. 2007. pp. 7-14 (Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007).

@inproceedings{b23bd39e78ec4eefa480753b96402743,

title = "Toward open-set text-independent speaker identification in tactical communications",

abstract = "We present the design and implementation of an open-set textindependent speaker identification system using genetic Learning Classifier Systems (LCS). We examine the use of this system in a real-number problem domain, where there is strong interest in its application to tactical communications. We investigate different encoding methods for representing real-number knowledge and study the efficacy of each method for speaker identification. We also identify several difficulties in solving the speaker identification problems with LCS and introduce new approaches to resolve the difficulties. Experimental results show that our system successfully learns 200 voice features at accuracies of 90 % to 100 % and 15,000 features to more than 80% for the closed-set problem, which is considered a strong result in the speaker identification community. The open-set capability is also comparable to existing numeric-based methods.",

keywords = "Classifier systems, Genetic algorithms, Language and speech, Machine learning",

author = "Wolf, {Matt B.} and Park, {Won Kyung} and Oh, {Jae C.} and Blowers, {Misty K.}",

year = "2007",

doi = "10.1109/CISDA.2007.368129",

language = "English (US)",

isbn = "1424407001",

series = "Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007",

pages = "7--14",

booktitle = "Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007",

note = "2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007 ; Conference date: 01-04-2007 Through 05-04-2007",

}

TY - GEN

T1 - Toward open-set text-independent speaker identification in tactical communications

AU - Wolf, Matt B.

AU - Park, Won Kyung

AU - Oh, Jae C.

AU - Blowers, Misty K.

PY - 2007

Y1 - 2007

N2 - We present the design and implementation of an open-set textindependent speaker identification system using genetic Learning Classifier Systems (LCS). We examine the use of this system in a real-number problem domain, where there is strong interest in its application to tactical communications. We investigate different encoding methods for representing real-number knowledge and study the efficacy of each method for speaker identification. We also identify several difficulties in solving the speaker identification problems with LCS and introduce new approaches to resolve the difficulties. Experimental results show that our system successfully learns 200 voice features at accuracies of 90 % to 100 % and 15,000 features to more than 80% for the closed-set problem, which is considered a strong result in the speaker identification community. The open-set capability is also comparable to existing numeric-based methods.

AB - We present the design and implementation of an open-set textindependent speaker identification system using genetic Learning Classifier Systems (LCS). We examine the use of this system in a real-number problem domain, where there is strong interest in its application to tactical communications. We investigate different encoding methods for representing real-number knowledge and study the efficacy of each method for speaker identification. We also identify several difficulties in solving the speaker identification problems with LCS and introduce new approaches to resolve the difficulties. Experimental results show that our system successfully learns 200 voice features at accuracies of 90 % to 100 % and 15,000 features to more than 80% for the closed-set problem, which is considered a strong result in the speaker identification community. The open-set capability is also comparable to existing numeric-based methods.

KW - Classifier systems

KW - Genetic algorithms

KW - Language and speech

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=34548710166&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548710166&partnerID=8YFLogxK

U2 - 10.1109/CISDA.2007.368129

DO - 10.1109/CISDA.2007.368129

M3 - Conference contribution

AN - SCOPUS:34548710166

SN - 1424407001

SN - 9781424407002

T3 - Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007

SP - 7

EP - 14

BT - Proceedings of the 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007

T2 - 2007 IEEE Symposium on Computational Intelligence in Security and Defense Applications, CISDA 2007

Y2 - 1 April 2007 through 5 April 2007

ER -

Toward open-set text-independent speaker identification in tactical communications

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this