UNLIMITED TEXT-TO-SPEECH SYSTEM: DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE.

Jared Bernstein, David Pisoni

Research output: Chapter in Book/Report/Conference proceedingChapter

16 Citations (Scopus)

Abstract

An English text-to-speech system based on MITalk has been developed which runs in real time on microprocessor hardware and in main memory. Certain parts of the MITalk algorithm have been omitted and replaced in the TSI system by an 1100 word dictionary and a simple ″hat and declination″ FO routine. Results from intelligibility and comprehension tests identical to those used in evaluating MITalk indicate that for common paragraphic text listener performance on TSI's speech is comparable to performance on MITalk speech. Analysis of the results suggests that accurate acoustic realization of segmental information is the crucial factor in intelligibility, with simple FO patterns and conventional letter-to-phoneme conversion adequate for many purposes.

Original languageEnglish
Title of host publicationRecord - IEEE International Conference on Acoustics, Speech & Signal Processing
PublisherIEEE
Pages576-579
Number of pages4
Volume2
StatePublished - 1980
Externally publishedYes
EventUnknown conference - Denver, CO, USA
Duration: Apr 9 1980Apr 11 1980

Other

OtherUnknown conference
CityDenver, CO, USA
Period4/9/804/11/80

Fingerprint

Microprocessor chips
Glossaries
Computer hardware
Acoustics
Data storage equipment

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Bernstein, J., & Pisoni, D. (1980). UNLIMITED TEXT-TO-SPEECH SYSTEM: DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE. In Record - IEEE International Conference on Acoustics, Speech & Signal Processing (Vol. 2, pp. 576-579). IEEE.

UNLIMITED TEXT-TO-SPEECH SYSTEM : DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE. / Bernstein, Jared; Pisoni, David.

Record - IEEE International Conference on Acoustics, Speech & Signal Processing. Vol. 2 IEEE, 1980. p. 576-579.

Research output: Chapter in Book/Report/Conference proceedingChapter

Bernstein, J & Pisoni, D 1980, UNLIMITED TEXT-TO-SPEECH SYSTEM: DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE. in Record - IEEE International Conference on Acoustics, Speech & Signal Processing. vol. 2, IEEE, pp. 576-579, Unknown conference, Denver, CO, USA, 4/9/80.
Bernstein J, Pisoni D. UNLIMITED TEXT-TO-SPEECH SYSTEM: DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE. In Record - IEEE International Conference on Acoustics, Speech & Signal Processing. Vol. 2. IEEE. 1980. p. 576-579
Bernstein, Jared ; Pisoni, David. / UNLIMITED TEXT-TO-SPEECH SYSTEM : DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE. Record - IEEE International Conference on Acoustics, Speech & Signal Processing. Vol. 2 IEEE, 1980. pp. 576-579
@inbook{b55478a8e7b04f1b85b0345479b43ebf,
title = "UNLIMITED TEXT-TO-SPEECH SYSTEM: DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE.",
abstract = "An English text-to-speech system based on MITalk has been developed which runs in real time on microprocessor hardware and in main memory. Certain parts of the MITalk algorithm have been omitted and replaced in the TSI system by an 1100 word dictionary and a simple ″hat and declination″ FO routine. Results from intelligibility and comprehension tests identical to those used in evaluating MITalk indicate that for common paragraphic text listener performance on TSI's speech is comparable to performance on MITalk speech. Analysis of the results suggests that accurate acoustic realization of segmental information is the crucial factor in intelligibility, with simple FO patterns and conventional letter-to-phoneme conversion adequate for many purposes.",
author = "Jared Bernstein and David Pisoni",
year = "1980",
language = "English",
volume = "2",
pages = "576--579",
booktitle = "Record - IEEE International Conference on Acoustics, Speech & Signal Processing",
publisher = "IEEE",

}

TY - CHAP

T1 - UNLIMITED TEXT-TO-SPEECH SYSTEM

T2 - DESCRIPTION AND EVALUATION OF A MICROPROCESSOR BASED DEVICE.

AU - Bernstein, Jared

AU - Pisoni, David

PY - 1980

Y1 - 1980

N2 - An English text-to-speech system based on MITalk has been developed which runs in real time on microprocessor hardware and in main memory. Certain parts of the MITalk algorithm have been omitted and replaced in the TSI system by an 1100 word dictionary and a simple ″hat and declination″ FO routine. Results from intelligibility and comprehension tests identical to those used in evaluating MITalk indicate that for common paragraphic text listener performance on TSI's speech is comparable to performance on MITalk speech. Analysis of the results suggests that accurate acoustic realization of segmental information is the crucial factor in intelligibility, with simple FO patterns and conventional letter-to-phoneme conversion adequate for many purposes.

AB - An English text-to-speech system based on MITalk has been developed which runs in real time on microprocessor hardware and in main memory. Certain parts of the MITalk algorithm have been omitted and replaced in the TSI system by an 1100 word dictionary and a simple ″hat and declination″ FO routine. Results from intelligibility and comprehension tests identical to those used in evaluating MITalk indicate that for common paragraphic text listener performance on TSI's speech is comparable to performance on MITalk speech. Analysis of the results suggests that accurate acoustic realization of segmental information is the crucial factor in intelligibility, with simple FO patterns and conventional letter-to-phoneme conversion adequate for many purposes.

UR - http://www.scopus.com/inward/record.url?scp=0019210598&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0019210598&partnerID=8YFLogxK

M3 - Chapter

AN - SCOPUS:0019210598

VL - 2

SP - 576

EP - 579

BT - Record - IEEE International Conference on Acoustics, Speech & Signal Processing

PB - IEEE

ER -