Translator Disclaimer
13 November 2003 Estimating speaker scale factors from vowels
Author Affiliations +
In previous works, Umesh et al, demonstrated that phonetically similar vowels spoken by different individuals are related by a simple translation in a universal warped spectral representation. They experimentally derived this function and called it the “speech-scale”. We present further experimental evidence, based on a large data set, validating the speech-scale. We also estimate speaker-specific scale factors based on the speech-scale, and we present a vowel classification experiment, which demonstrates a significant performance improvement through a normalization based on the speech-scale. The results we present are based on formant estimates of vowels in a Western Michigan vowel database.
© (2003) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Douglas J. Nelson, David C. Smith, Srinivasan Umesh, and Leon Cohen "Estimating speaker scale factors from vowels", Proc. SPIE 5207, Wavelets: Applications in Signal and Image Processing X, (13 November 2003);


Biometric speaker classification
Proceedings of SPIE (October 13 2000)
New basis function distortion invariant detection filters
Proceedings of SPIE (September 29 1994)
The speech scale, the Mel scale, and the tube model...
Proceedings of SPIE (December 06 2002)
Linear distribution of signals
Proceedings of SPIE (October 26 2004)
Further experimental results on the speech-hearing connection
Proceedings of SPIE (December 05 2001)

Back to Top