Fact · Accessibility and health
2hours
In the 2007 study, four groups of speakers each did a speech recognition task for 2 hours, with voice samples recorded before and after.
Checked against its source on October 16, 2026
In the source's own words
“Four different subject groups carried out a speech recognition task for 2 hours.”
Voice fatigue and use of speech recognition: a study of voice quality ratings, Christel de Bruijn and Sandra Whiteside, Proceedings of Interspeech 2007 (ISCA), 2007, 2. Methodology.
Open the original (DOI 10.21437/Interspeech.2007-384)- What it measures
- The study compared discrete recognition, which needs a pause between words, with continuous recognition, and speakers with high and low vocal loads.
- What it does not show
- Small study with lab recordings and a listeners' panel. It measures voice quality after one session, not long-term damage.