SuperLectures.com

UTILIZING GLOTTAL SOURCE PULSE LIBRARY FOR GENERATING IMPROVED EXCITATION SIGNAL FOR HMM-BASED SPEECH SYNTHESIS

Full Paper at IEEE Xplore

Speech Synthesis

Presented by: Tuomo Raitio, Author(s): Tuomo Raitio, Aalto University, Finland; Antti Suni, University of Helsinki, Finland; Hannu Pulakka, Aalto University, Finland; Martti Vainio, University of Helsinki, Finland; Paavo Alku, Aalto University, Finland

This paper describes a source modeling method for hidden Markov model (HMM) based speech synthesis for improved naturalness. A speech corpus is first decomposed into the glottal source signal and the model of the vocal tract filter using glottal inverse filtering, and parametrized into excitation and spectral features. Additionally, a library of glottal source pulses is extracted from the estimated voice source signal. In the synthesis stage, the excitation signal is generated by selecting appropriate pulses from the library according to the target cost of the excitation features and a concatenation cost between adjacent glottal source pulses. Finally, speech is synthesized by filtering the excitation signal by the vocal tract filter. Experiments show that the naturalness of the synthetic speech is better or equal, and speaker similarity is better, compared to a system using only single glottal source pulse.


  Speech Transcript

|

  Slides

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

0:00:16

  1. slide

0:00:47

  2. slide

0:00:53

  3. slide

0:01:33

  4. slide

0:02:19

  5. slide

0:02:35

  6. slide

0:02:50

  7. slide

0:03:55

  8. slide

0:04:46

  9. slide

0:05:36

 10. slide

0:05:56

 11. slide

0:06:30

 12. slide

0:07:10

 13. slide

0:08:08

 14. slide

0:08:24

 15. slide

0:09:39

 16. slide

0:10:37

 17. slide

0:11:17

 18. slide

0:12:01

 19. slide

0:12:28

 20. slide

0:15:47

     9. slide

0:15:54

     8. slide

  Comments

Please sign in to post your comment!

  Lecture Information

Recorded: 2011-05-26 14:05 - 14:25, Panorama
Added: 15. 6. 2011 18:16
Number of views: 48
Video resolution: 1024x576 px, 512x288 px
Video length: 0:17:58
Audio track: MP3 [6.06 MB], 0:17:58