Comparison of Two Methods of Voice Activity Detection in Field Studies PurposeTo evaluate and compare the performance of 2 methods of voice activity detection (neck-attached accelerometer vs. binaural recordings) in field studies in environments where voice activity normally occurs.MethodA group of 11 healthy adults wore recording equipment during their lunch break. We used binary classification to analyze the results from the ... Research Note
Research Note  |   December 01, 2009
Comparison of Two Methods of Voice Activity Detection in Field Studies
 
Author Affiliations & Notes
  • Fredric Lindstrom
    University of Gothenburg, Sweden
  • Keni Ren
    Umeå University, Umeå, Sweden
  • Haibo Li
    Umeå University, Umeå, Sweden
  • Kerstin Persson Waye
    University of Gothenburg
  • Contact author: Fredric Lindstrom, Gothenburg University, Department of Public Health and Community Medicine, Box 414, Gothenburg SE-405 30, Sweden. E-mail: fredric.lindstrom@amm.gu.se.
Article Information
Speech, Voice & Prosody / Speech
Research Note   |   December 01, 2009
Comparison of Two Methods of Voice Activity Detection in Field Studies
Journal of Speech, Language, and Hearing Research, December 2009, Vol. 52, 1658-1663. doi:10.1044/1092-4388(2009/08-0175)
History: Received August 24, 2008 , Accepted March 12, 2009
 
Journal of Speech, Language, and Hearing Research, December 2009, Vol. 52, 1658-1663. doi:10.1044/1092-4388(2009/08-0175)
History: Received August 24, 2008; Accepted March 12, 2009
Web of Science® Times Cited: 6

PurposeTo evaluate and compare the performance of 2 methods of voice activity detection (neck-attached accelerometer vs. binaural recordings) in field studies in environments where voice activity normally occurs.

MethodA group of 11 healthy adults wore recording equipment during their lunch break. We used binary classification to analyze the results from the 2 methods. The output was compared to a gold standard, obtained through listening tests, and the probability for sensitivity (Ps) and false positive (Pf) was rated. The binary classifiers were set for consistent sensitivity of 99%; thus, the lower false positive rate would indicate the method with the better performance.

ResultsThe neck-attached accelerometer (Pf = 0.5%) performed significantly (p < .001) better than the binaural method (Pf = 7%).

ConclusionThe neck-attached accelerometer is more suitable than the binaural method for voice assessments in environments where people are speaking in close proximity to each other and where the signal-to-noise ratio is moderate to low.

Acknowledgment
This study was supported by the Swedish Research Council (Vetenskapsrådet [VR]).
Order a Subscription
Pay Per View
Entire Journal of Speech, Language, and Hearing Research content & archive
24-hour access
This Article
24-hour access