SuperLectures.com

A SYMMETRIZATION OF THE SUBSPACE GAUSSIAN MIXTURE MODEL

Acoustic Modeling

Full Paper at IEEE Xplore

Presented by: Daniel Povey, Author(s): Daniel Povey, Microsoft Corporation, United States; Martin Karafiát, Brno University of Technology, Czech Republic; Arnab Ghoshal, University of Saarland, Czech Republic; Petr Schwarz, Brno University of Technology, Czech Republic

Last year we introduced the Subspace Gaussian Mixture Model (SGMM), and we demonstrated Word Error Rate improvements on a fairly small-scale task. Here we describe an extension to the SGMM, which we call the symmetric SGMM. It makes the model fully symmetric between the ``speech-state vectors'' and ``speaker vectors'' by making the mixture weights depend on the speaker as well as the speech state. We find that the symmetric SGMM can give a very worthwhile improvement over the previously described model. We will also describe some larger-scale experiments with the SGMM, and report on progress toward releasing open-source software that supports SGMMs.


  Speech Transcript

|

  Slides

Enlarge the slide | Show all slides in a pop-up window

0:00:16

  1. slide

0:00:41

  2. slide

0:01:41

  3. slide

0:02:07

  4. slide

0:02:36

  5. slide

0:03:58

  6. slide

0:05:37

  7. slide

0:06:55

  8. slide

0:09:09

  9. slide

0:09:21

     8. slide

0:10:05

 10. slide

0:11:01

 11. slide

0:12:18

 12. slide

0:12:41

 13. slide

0:13:42

 14. slide

0:14:07

 15. slide

0:14:33

 16. slide

0:14:48

 17. slide

0:15:01

 18. slide

0:15:37

 19. slide

  Comments

Please sign in to post your comment!

  Lecture Information

Recorded: 2011-05-25 15:05 - 15:25, Panorama
Added: 15. 6. 2011 16:17
Number of views: 182
Video resolution: 1024x576 px, 512x288 px
Video length: 0:19:39
Audio track: MP3 [6.64 MB], 0:19:39