Research Article Open Access

Comparison of Speech Features on the Speech Recognition Task

Iosif Mporas1, Todor Ganchev1, Mihalis Siafarikas1 and Nikos Fakotakis1
  • 1 ,
Journal of Computer Science
Volume 3 No. 8, 2007, 608-616

DOI: https://doi.org/10.3844/jcssp.2007.608.616

Submitted On: 21 August 2007 Published On: 31 December 2007

How to Cite: Mporas, I., Ganchev, T., Siafarikas, M. & Fakotakis, N. (2007). Comparison of Speech Features on the Speech Recognition Task. Journal of Computer Science, 3(8), 608-616. https://doi.org/10.3844/jcssp.2007.608.616

Abstract

In the present work we overview some recently proposed discrete Fourier transform (DFT)- and discrete wavelet packet transform (DWPT)-based speech parameterization methods and evaluate their performance on the speech recognition task. Specifically, in order to assess the practical value of these less studied speech parameterization methods, we evaluate them in a common experimental setup and compare their performance against traditional techniques, such as the Mel-frequency cepstral coefficients (MFCC) and perceptual linear predictive (PLP) cepstral coefficients which presently dominate the speech recognition field. In particular, utilizing the well established TIMIT speech corpus and employing the Sphinx-III speech recognizer, we present comparative results of 8 different speech parameterization techniques.

  • 1,235 Views
  • 2,670 Downloads
  • 18 Citations

Download

Keywords

  • Speech parameterization
  • speech recognition
  • wavelet packets