Estimating Speaker Scale Factors from Vowels

D. J. Nelson, D. C. Smith, S. Umesh, L. Cohen

Research output: Contribution to journalConference articlepeer-review

2 Scopus citations

Abstract

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.

Original languageEnglish
Pages (from-to)794-800
Number of pages7
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5207
Issue number2
DOIs
StatePublished - 2003
Externally publishedYes
EventWavelets: Applications in Signal and Image Processing X - San Diego, CA, United States
Duration: Aug 4 2003Aug 8 2003

Keywords

  • Cross-spectrum
  • Formant estimation
  • Speaker scale
  • Spectral correlation
  • Spectral estimation

Fingerprint

Dive into the research topics of 'Estimating Speaker Scale Factors from Vowels'. Together they form a unique fingerprint.

Cite this