Speaker Verification Using Adaptive Dictionaries in Non-negative Spectrogram Deconvolution

  • Szymon DrgasEmail author
  • Tuomas Virtanen
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9237)


This article presents a new method for speaker verification, which is based on the non-negative matrix deconvolution (NMD) of the magnitude spectrogram of an observed utterance. In contrast to typical methods known from the literature, which are based on the assumption that the desired signal dominates (for example GMM-UBM, joint factor analysis, i-vectors), compositional models such as NMD describe a recording as a non-negative combination of latent components. The proposed model represents a spectrogram of a signal as a sum of spectro-temporal patterns that span durations of order about 150 ms, while many state of the art automatic speaker recognition systems model a probability distribution of features extracted from much shorter excerpts of speech signal (about 50 ms). Longer patterns carry information about dynamical aspects of modeled signal, for example information about accent and articulation. We use a parametric dictionary in the NMD and the parameters of the dictionary carry information about the speakers’ identity. The experiments performed on the CHiME corpus show that with the proposed approach achieves equal error rate comparable to an i-vector based system.


Compositional Model Equal Error Rate Speaker Recognition Speaker Verification Speaker Identification 
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.
    Kinnunen, T., Li, H.: An overview of text-independent speaker recognition: From features to supervectors. Speech Commun. 52, 12–40 (2010)CrossRefGoogle Scholar
  2. 2.
    McLaren, M., Scheffer, N., Ferrer, L., Lei, Y.: Effective use of DCTs for contextualizing features for speaker. In: IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP) (2014)Google Scholar
  3. 3.
    Kinnunen, T., Wei, C., Koh, E., Wang, L., Li, H., Chng, E.S.: Temporal discrete cosine transform: Towards longer term temporal features for speaker verification. In: Proceedings of the Fifth International Symposium on Chinese Spoken Language Processing (ISCSLP 2006), Singapore, pp. 547–558 (2006)Google Scholar
  4. 4.
    Virtanen, T., Gemmeke, J.F., Raj, B., Smaragdis, P.: Compositional models for audio processing. IEEE Signal Process. Mag. 32(2), 125–144 (2015)CrossRefGoogle Scholar
  5. 5.
    Saeidi, R., Hurmaleinen, A., Virtanen, T., van Leeuwen, D.A.: Exemplar-based sparse representation and sparse discrimination for noise robust speaker identification. In: Odyssey 2012: The Speaker and Langueage Recognition Workshop (2012)Google Scholar
  6. 6.
    Joder, C., Schuller, B.: Exploring nonnegative matrix factorization for audio classification: application to speaker recognition. In: Proceedings of Speech Communication; 10, ITG Symposium, VDE, 1–4 (2012)Google Scholar
  7. 7.
    Smaragdis, P.: Convolutive speech bases and their application to supervised speech separation. IEEE Trans. Speech and Audio Process. 15(1), 1–12 (2005)CrossRefGoogle Scholar
  8. 8.
    Hurmalainen, A., Virtanen, T.: Learning state labels for sparse classification of speech with matrix deconvolution. IEEE Workshop Autom. Speech Recogn. Underst. (ASRU) 2013, 168–173 (2013)Google Scholar
  9. 9.
    Ding, C., Li, T., Jordan, M.: Convex and semi-nonnegative matrix factorizations. IEEE Trans. Pattern Anal. Mach. Intell. 32, 45–55 (2010)CrossRefGoogle Scholar
  10. 10.
    Christensen, H., Barker, J., Ma, N., Green, P.D.: The CHiME corpus: a resource and a challenge for computational hearing in multisource environments. In: INTERSPEECH, pp. 1918–1921 (2010)Google Scholar
  11. 11.
    Sadjadi, S.O., Slaney, M., Heck, L.: MSR identity toolbox v1.0: A MATLAB toolbox for speaker-recognition research. Speech and Language Processing Technical Committee Newsletter (2013)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.Chair of Control and Systems EngineeringPoznan University of TechnologyPoznańPoland
  2. 2.Department of Signal ProcessingTampere University of TechnologyTampereFinland

Personalised recommendations