advanced
Journal Information
Journal Information
   Description
   Editorial Board
   Guide for Authors
   Ordering

Contents Services
Contents Services
   Regular Issues
   Special Issues
   Authors Index

Links
Links
   FEI STU Bratislava
   SAS Bratislava

   Feedback

[2, 2018] 

Journal of Electrical Engineering, Vol 69, 2 (2018) 138-147 DOI: 10.2478/jee-2018-0017

Evaluation of speaker de-identification based on voice gender and age conversion

Jiří Přibil – Anna Přibilová – Jindřich Matoušek

   Two basic tasks are covered in this paper. The first one consists in the design and practical testing of a new method for voice de-identification that changes the apparent age and/or gender of a speaker by multi-segmental frequency scale transformation combined with prosody modification. The second task is aimed at verification of applicability of a classifier based on Gaussian mixture models (GMM) to detect the original Czech and Slovak speakers after applied voice de-identification. The performed experiments confirm functionality of the developed gender and age conversion for all selected types of de-identification which can be objectively evaluated by the GMM-based open-set classifier. The original speaker detection accuracy was compared also for sentences uttered by German and English speakers showing language independence of the proposed method.

Keywords: "GMM classifier, spectral and prosodic features of speech, speaker gender and age classification


[full-paper]


© 1997-2019  FEI STU Bratislava