Voice Fatique and use of Speech Recognition: Acoustic and Perceptual Measures

Christel de Bruijn, Sandra Whiteside

In the 1990s, after decades of research, automatic speech recognition (ASR) software finally came within reach of the general consumer market, and was no longer confined to research laboratories and purpose-built systems of large companies. However, as is often the case with the introduction of new technologies, reports of associated voice problems followed soon after. The reports appeared in various media, ranging from computer magazines, mailing lists for users of speech recognition software to the Bulletin of the Royal College for Speech and Language Therapists (UK).
The problems and symptoms reported varied in severity from a dry throat to being unable to speak for more than half an hour a day. So far, however, only two formal studies have investigated these claims. The aim of the study presented here was to investigate the effects of speech recognition on voice in an experimental setting. Two main variables have been taken into consideration in this study: type of ASR software and the typical vocal load of a user. Subjects with either a high or low vocal load undertook a 2-hour dictation task using either discrete or continuous ASR software. Sustained vowels (/A/,/i/,/u/) were recorded before and after the task for acoustic analysis and perceptual analysis by a panel of speech and language therapists. One of the most consistent results was a rise of F0, as well as (perceived) pitch, a common finding in voice fatigue studies. The results are interpreted as part of a hyperfunctional mechanism countering early signs of voice fatigue.

Christel de Bruijn
Division of Speech and Language Therapy University of Central England

Perry Barr
BirminghamB42 2SU UK
phone: +44 (0)121 331 5720 fax: +44 (0)121 331 6207
e-mail: christel.debruijn@uce.ac.uk

Sandra Whiteside
Department of Human Communication Sciences
Universityof Sheffield
31 Claremont Crescent
Sheffield S10 2TA
phone: +44 (0)114 222 2447 fax: +44 (0)114 273 0547
e-mail: s.whiteside@sheffield.ac.uk