Cognitive Computation

, Volume 5, Issue 4, pp 517–525

Global Selection of Features for Nonlinear Dynamics Characterization of Emotional Speech

  • Patricia Henríquez Rodríguez
  • Jesús B. Alonso Hernández
  • Miguel A. Ferrer Ballester
  • Carlos M. Travieso González
  • Juan R. Orozco-Arroyave
Article

DOI: 10.1007/s12559-012-9157-0

Cite this article as:
Henríquez Rodríguez, P., Alonso Hernández, J.B., Ferrer Ballester, M.A. et al. Cogn Comput (2013) 5: 517. doi:10.1007/s12559-012-9157-0

Abstract

This paper proposes the application of measures based on nonlinear dynamics for emotional speech characterization. Measures such as mutual information, dimension correlation, entropy correlation, Shannon’s entropy, Lempel–Ziv complexity and Hurst exponent are extracted from the samples of a database of emotional speech. Then, summary statistics such as mean, standard deviation, skewness and kurtosis are applied on the extracted measures. Experiments were conducted on the Berlin emotional speech database for a three-class problem (neutral, fear and anger as emotional states). Feature selection is accomplished and a methodology is proposed to find the best features. In order to evaluate the discrimination ability of the selected features, a neural network classifier is used. The global success rate is 93.78 ± 3.18 %.

Keywords

Nonlinear dynamic Emotional speech Forward floating feature selection Neural networks 

Copyright information

© Springer Science+Business Media, LLC 2012

Authors and Affiliations

  • Patricia Henríquez Rodríguez
    • 1
  • Jesús B. Alonso Hernández
    • 1
  • Miguel A. Ferrer Ballester
    • 1
  • Carlos M. Travieso González
    • 1
  • Juan R. Orozco-Arroyave
    • 2
  1. 1.Instituto Universitario para el Desarrollo Tecnológico e Innovación en Comunicaciones (IDeTIC)Universidad de Las Palmas de Gran CanariaLas Palmas de Gran CanariaSpain
  2. 2.Departamento de Ingeniería ElectrónicaUniversidad de Antioquia. GEPAR and GITA Research GroupsMedellínColombia