Recognition of speech spectrograms

B. G. Greene, David Pisoni, T. D. Carrell

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

The performance of eight naive observers in learning to identify speech spectrograms was studied over a 2-month period. Single tokens from a 50-word phonetically balanced (PB) list were recorded by several talkers and displayed on a Spectraphonics Speech Spectrographic Display system. Identification testing occurred immediately after daily training sessions. After approximately 20 hr of training, naive subjects correctly identified the 50 PB words from a single talker over 95% of the time. Generalization tests with the same words were then carried out with different tokens from the original talker, new tokens from another male talker, a female talker, and finally, a synthetic talker. The generalization results for these talkers showed recognition performance at 91%, 76%, 76%, and 48%, respectively.

Original languageEnglish
Pages (from-to)32-43
Number of pages12
JournalJournal of the Acoustical Society of America
Volume76
Issue number1
StatePublished - Jul 1984

Fingerprint

spectrograms
education
display devices
lists
learning
Talkers

ASJC Scopus subject areas

  • Acoustics and Ultrasonics

Cite this

Recognition of speech spectrograms. / Greene, B. G.; Pisoni, David; Carrell, T. D.

In: Journal of the Acoustical Society of America, Vol. 76, No. 1, 07.1984, p. 32-43.

Research output: Contribution to journalArticle

Greene, BG, Pisoni, D & Carrell, TD 1984, 'Recognition of speech spectrograms', Journal of the Acoustical Society of America, vol. 76, no. 1, pp. 32-43.
Greene, B. G. ; Pisoni, David ; Carrell, T. D. / Recognition of speech spectrograms. In: Journal of the Acoustical Society of America. 1984 ; Vol. 76, No. 1. pp. 32-43.
@article{cbe4dc8260944d27a14b32152583e52e,
title = "Recognition of speech spectrograms",
abstract = "The performance of eight naive observers in learning to identify speech spectrograms was studied over a 2-month period. Single tokens from a 50-word phonetically balanced (PB) list were recorded by several talkers and displayed on a Spectraphonics Speech Spectrographic Display system. Identification testing occurred immediately after daily training sessions. After approximately 20 hr of training, naive subjects correctly identified the 50 PB words from a single talker over 95{\%} of the time. Generalization tests with the same words were then carried out with different tokens from the original talker, new tokens from another male talker, a female talker, and finally, a synthetic talker. The generalization results for these talkers showed recognition performance at 91{\%}, 76{\%}, 76{\%}, and 48{\%}, respectively.",
author = "Greene, {B. G.} and David Pisoni and Carrell, {T. D.}",
year = "1984",
month = "7",
language = "English",
volume = "76",
pages = "32--43",
journal = "Journal of the Acoustical Society of America",
issn = "0001-4966",
publisher = "Acoustical Society of America",
number = "1",

}

TY - JOUR

T1 - Recognition of speech spectrograms

AU - Greene, B. G.

AU - Pisoni, David

AU - Carrell, T. D.

PY - 1984/7

Y1 - 1984/7

N2 - The performance of eight naive observers in learning to identify speech spectrograms was studied over a 2-month period. Single tokens from a 50-word phonetically balanced (PB) list were recorded by several talkers and displayed on a Spectraphonics Speech Spectrographic Display system. Identification testing occurred immediately after daily training sessions. After approximately 20 hr of training, naive subjects correctly identified the 50 PB words from a single talker over 95% of the time. Generalization tests with the same words were then carried out with different tokens from the original talker, new tokens from another male talker, a female talker, and finally, a synthetic talker. The generalization results for these talkers showed recognition performance at 91%, 76%, 76%, and 48%, respectively.

AB - The performance of eight naive observers in learning to identify speech spectrograms was studied over a 2-month period. Single tokens from a 50-word phonetically balanced (PB) list were recorded by several talkers and displayed on a Spectraphonics Speech Spectrographic Display system. Identification testing occurred immediately after daily training sessions. After approximately 20 hr of training, naive subjects correctly identified the 50 PB words from a single talker over 95% of the time. Generalization tests with the same words were then carried out with different tokens from the original talker, new tokens from another male talker, a female talker, and finally, a synthetic talker. The generalization results for these talkers showed recognition performance at 91%, 76%, 76%, and 48%, respectively.

UR - http://www.scopus.com/inward/record.url?scp=0021461482&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0021461482&partnerID=8YFLogxK

M3 - Article

VL - 76

SP - 32

EP - 43

JO - Journal of the Acoustical Society of America

JF - Journal of the Acoustical Society of America

SN - 0001-4966

IS - 1

ER -