KPCA vs. PCA study for an age classification of speakers

  • Authors:

  • Cristina Muñoz-Mulas;Rafael Martínez-Olalla;Pedro Gómez-Vilda;Elmar W. Lang;Agustín Álvarez-Marquina;Luis Miguel Mazaira-Fernández;Víctor Nieto-Lluis

  • Affiliations:

  • Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid;Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid;Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid;Computational Intelligence and Machine Learning Group, Institut of Biophysik, Universität of Regensburg, Regensburg, Germany;Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid;Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid;Departamento de Arquitectura de Tecnologías y Sistemas Informáticos, Universidad Politécnica de Madrid, Madrid

  • Venue:
  • NOLISP'11 Proceedings of the 5th international conference on Advances in nonlinear speech processing
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Kernel-PCA and PCA techniques are compared in the task of age and gender separation. A feature extraction process that discriminates between vocal tract and glottal source is implemented. The reason why speech is processed in that way is because vocal tract length and resonant characteristics are related to gender and age and there is also a great relationship between glottal source and age and gender. The obtained features are then processed with PCA and kernel-PCA techniques. The results show that gender and age separation is possible and that kernel-PCA (especially with RBF kernel) clearly outperforms classical PCA or no preprocessing features.