Affect-Insensitive Speaker Recognition by Feature Variety Training

  • Dongdong Li
  • Yingchun Yang
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4738)


A great deal of inner variabilities such as emotion and stress are largely missing from traditional speaker recognition system. The direct result is that the recognition system is easily disturbed when the enrollment and the authentication are made under different emotional state. Reynolds [1] proposed a new normalization technique called feature mapping. This technique achieved big successes in channel robust speaker verification. We extend the mapping idea to develop a feature variety training approach for affective-insensitive speaker recognition.


Gaussian Mixture Model Speaker Recognition Feature Transformation Emotion Speech Target Emotion 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Reynolds, D.A.: Channel robust speaker verification via feature mapping. In: ICASSP 2003, vol. 2, pp. 53–56 (2003)Google Scholar
  2. 2.
    Wu, T., Yang, Y.C., Wu, Z.H., Li, D.D., MASC,: A Speech Corpus in Mandarin for Emotion Analysis and Affective Speaker Recognition, The IEEE Odyssey, 1–59 (2006)Google Scholar
  3. 3.
    Bonastre, J.F., Wils, F., Meignier, S.: ALIZE, a free toolkit for speaker recognition. In: IEEE International Conference on Acoustics, Speech, and Signal Processing. In (ICASSP 2005), March 18-23, vol. 1, pp. 737–740 (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Dongdong Li
    • 1
  • Yingchun Yang
    • 1
  1. 1.Department of Computer Science and Technology, Zhejiang University, Hangzhou, 310027P.R. China

Personalised recommendations