SuperLectures.com

INTRA-SESSION VARIABILITY COMPENSATION AND A HYPOTHESIS GENERATION AND SELECTION STRATEGY FOR SPEAKER SEGMENTATION

Miscellaneous Speaker Identification

Full Paper at IEEE Xplore

Přednášející: Carlos Vaquero, Autoři: Carlos Vaquero, Alfonso Ortega, Eduardo Lleida, University of Zaragoza, Spain

This paper addresses the problem of speaker segmentation in two-speaker telephone conversations, using an eigenvoice based factor analysis approach. We present a set of improvements in the speaker segmentation system. First, we study two methods to compensate for intra-session variability, that is the variability present in a speaker during a single session. Secondly we propose a method to generate segmentation hypotheses that combined with a given confidence measure, enables the selection of correct hypotheses improving the overall segmentation performance. The proposed improvements are evaluated on the NIST Speaker Recognition Evaluation 2008 summed channel test condition, obtaining 28% relative improvement in terms of speaker segmentation error.


  Přepis řeči

|

  Slajdy

Zvětšit slajd | Zobrazit všechny slajdy

0:00:16

  1. slajd

0:00:46

  2. slajd

0:01:23

  3. slajd

0:02:10

  4. slajd

0:03:00

  5. slajd

0:04:50

  6. slajd

0:05:46

  7. slajd

0:06:24

  8. slajd

0:07:01

  9. slajd

0:07:29

 10. slajd

0:08:07

 11. slajd

0:09:09

 12. slajd

0:09:39

 13. slajd

0:10:46

 14. slajd

0:11:46

 15. slajd

0:12:33

 16. slajd

0:13:42

 17. slajd

0:14:26

 18. slajd

0:16:18

 19. slajd

  Komentáře

Please sign in to post your comment!

  Informace o přednášce

Nahráno: 2011-05-25 17:55 - 18:15, Panorama
Přidáno: 15. 6. 2011 16:58
Počet zhlédnutí: 13
Rozlišení videa: 1024x576 px, 512x288 px
Délka videa: 0:19:01
Audio stopa: MP3 [6.42 MB], 0:19:01