THE FOURTH ORDER CUMULANT OF SPEECH SIGNALS APPLIED TO PITCH ESTIMATION

H MAALEM, F MARIR

Résumé


In a number of speech applications, such as coding, synthesis or recognition, it is crucial to
make a reliable discrimination between voiced/unvoiced segments and accurately determine
the pitch period. The problem of an accurate estimation and decision in noisy condition
remains open; Higher-order statistics (H.O.S) have inherent properties that make them well
suited when dealing with a mixture of Gaussian and non-Gaussian processes.
This paper explores the fourth order cumulant using autoregressive model (AR(p)) and
presents a new algorithm for pitch detection of voiced sounds with and without colored
Gaussian noise and shows the superiority of the novel method over the classical methods such
as cepstral method.

Mots-clés


Pitch; Formants; Cumulants; Autoregressive model (AR); Cepstre

Texte intégral :

PDF

Références


- E. Nemer, R.A. Goubran and al "Voicing decision and pitch

estimation using third order statistics" Proc. ISCPAT-97,

San Diego, CA, USA, September 1997.

- Elias Nemer, and al "The fourth order cumulant of speech

signals with application to voice activity detection" IEEE

Trans. On signal Proc., vol. 42, N° 1, Jan. 1994, pp. 222-

- J. Mendel ,"Tutorial on higher- order statistics" Proc. IEEE,

vol 79, N°3, March 1991

- L. Kadri « Analyse spectrale du signal parole arabe

application aux enfants à audition déficiente » Mémoire de

magister, Université de Constantine 2000.

- J. Markel and al « Linear prediction of speech », Springer-

Verlag, 1976.

- J.W.A. Fackrell and al, "quadratic phase coupling detection

using H.O.S" IEE Coll. H.O.S. Savoy, Place, London, 22

May 1995


Renvois

  • Il n'y a présentement aucun renvoi.




Direction des Publications et de l'animation scientifique

Université des Frères Mentouri Constantine 1. Route Ain El-Bey. 25000. Algérie.