Interactive Clustering Techniques for Selecting Speaker-Independent Reference Templates for Isolated Word Recognition

Stephen E Levinson, Lawrence R. Rabiner, Aaron E. Rosenberg, Jay G. Wilpon

Research output: Contribution to journalArticle

Abstract

It is demonstrated that clustering can be a powerful tool for selecting reference templates for speaker-independent word recognition. We describe a set of clustering techniques specifically designed for this purpose. These interactive procedures identify coarse structure, fine structure, overlap of, and outliers from clusters. The techniques have been applied tp a large speech data base consisting of four repetitions of a 39 word vocabulary (the letters of the alphabet, the digits, and three auxiliary commands) spoken by 50 male and 50 female speakers. The results of the cluster analysis show that the data are highly structured containing large prominent clusters. Some statistics of the analysis and their significance are presented.

Original languageEnglish (US)
Pages (from-to)134-141
Number of pages8
JournalIEEE Transactions on Acoustics, Speech, and Signal Processing
Volume27
Issue number2
DOIs
StatePublished - Jan 1 1979
Externally publishedYes

Fingerprint

Cluster analysis
Statistics

ASJC Scopus subject areas

  • Signal Processing

Cite this

Interactive Clustering Techniques for Selecting Speaker-Independent Reference Templates for Isolated Word Recognition. / Levinson, Stephen E; Rabiner, Lawrence R.; Rosenberg, Aaron E.; Wilpon, Jay G.

In: IEEE Transactions on Acoustics, Speech, and Signal Processing, Vol. 27, No. 2, 01.01.1979, p. 134-141.

Research output: Contribution to journalArticle

@article{949a9baae7e14834b33d898b9b52890e,
title = "Interactive Clustering Techniques for Selecting Speaker-Independent Reference Templates for Isolated Word Recognition",
abstract = "It is demonstrated that clustering can be a powerful tool for selecting reference templates for speaker-independent word recognition. We describe a set of clustering techniques specifically designed for this purpose. These interactive procedures identify coarse structure, fine structure, overlap of, and outliers from clusters. The techniques have been applied tp a large speech data base consisting of four repetitions of a 39 word vocabulary (the letters of the alphabet, the digits, and three auxiliary commands) spoken by 50 male and 50 female speakers. The results of the cluster analysis show that the data are highly structured containing large prominent clusters. Some statistics of the analysis and their significance are presented.",
author = "Levinson, {Stephen E} and Rabiner, {Lawrence R.} and Rosenberg, {Aaron E.} and Wilpon, {Jay G.}",
year = "1979",
month = "1",
day = "1",
doi = "10.1109/TASSP.1979.1163222",
language = "English (US)",
volume = "27",
pages = "134--141",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Interactive Clustering Techniques for Selecting Speaker-Independent Reference Templates for Isolated Word Recognition

AU - Levinson, Stephen E

AU - Rabiner, Lawrence R.

AU - Rosenberg, Aaron E.

AU - Wilpon, Jay G.

PY - 1979/1/1

Y1 - 1979/1/1

N2 - It is demonstrated that clustering can be a powerful tool for selecting reference templates for speaker-independent word recognition. We describe a set of clustering techniques specifically designed for this purpose. These interactive procedures identify coarse structure, fine structure, overlap of, and outliers from clusters. The techniques have been applied tp a large speech data base consisting of four repetitions of a 39 word vocabulary (the letters of the alphabet, the digits, and three auxiliary commands) spoken by 50 male and 50 female speakers. The results of the cluster analysis show that the data are highly structured containing large prominent clusters. Some statistics of the analysis and their significance are presented.

AB - It is demonstrated that clustering can be a powerful tool for selecting reference templates for speaker-independent word recognition. We describe a set of clustering techniques specifically designed for this purpose. These interactive procedures identify coarse structure, fine structure, overlap of, and outliers from clusters. The techniques have been applied tp a large speech data base consisting of four repetitions of a 39 word vocabulary (the letters of the alphabet, the digits, and three auxiliary commands) spoken by 50 male and 50 female speakers. The results of the cluster analysis show that the data are highly structured containing large prominent clusters. Some statistics of the analysis and their significance are presented.

UR - http://www.scopus.com/inward/record.url?scp=0018455339&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0018455339&partnerID=8YFLogxK

U2 - 10.1109/TASSP.1979.1163222

DO - 10.1109/TASSP.1979.1163222

M3 - Article

VL - 27

SP - 134

EP - 141

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 2

ER -