| Paper: | SP-P3.10 | ||
| Session: | Topics in Speaker and Langauge Recognition | ||
| Time: | Tuesday, May 18, 15:30 - 17:30 | ||
| Presentation: | Poster | ||
| Topic: | Speech Processing: Speaker Recognition | ||
| Title: | BENEFITS OF PRIOR ACOUSTIC SEGMENTATION FOR AUTOMATIC SPEAKER SEGMENTATION | ||
| Authors: | Sylvain Meignier; Laboratoire Informatique d'Avignon (LIA) | ||
| Daniel Moraru; CLIPS-IMAG | |||
| Corinne Fredouille; Laboratoire Informatique d'Avignon (LIA) | |||
| Laurent Besacier; CLIPS-IMAG | |||
| Jean-François Bonastre; Laboratoire Informatique d'Avignon (LIA) | |||
| Abstract: | This paper investigates the interest of segmentation in acoustic macro classes (like gender or bandwidth) as a front-end processing for segmentation/diarization task. The impact of this prior acoustic segmentation is evaluated in terms of speaker diarization performance in the particular context of NIST RT’03 evaluation (done on HUB4 broadcast news corpora). Rarely discussed in the literature, this work shows that prior acoustic segmentation, in a similar way to automatic speech recognition task, may be very useful to speaker segmentation task. The experiments were conducted using two different kinds of speaker segmentation systems developed individually by the LIA and CLIPS laboratories in the framework of the ELISA consortium. For both systems, improvement was observed when combined with prior acoustic segmentation. However, a larger impact, in terms of performance, is observed on the ascending/HMM approach based LIA system compared to the speaker turn detection based CLIPS system. | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops