construction et évaluation d'un corpus audio-visuel pour la parole expressive vietnamienne.
Séminaire de M. Mac Dang Khoa, doctorant du Centre MICA - Date : vendredi 23 octobre, 16h00 - Lieu : Centre MICA

Intervenant :
M. Mac Dang Khoa
Doctorant en co-tutelle Centre MICA/ laboratoire LIG

Date : vendredi 23 octobre 2009, 16h00
Lieu : salle polyvalente, bâtiment C10, 4ème étage, Centre MICA
Interprète traducteur : le séminaire sera présenté en français

Abstract :
For a tonal language like Vietnamese, the acoustic parameters implied in the linguistic and affective functions of prosody (typically F0, intensity, timing) also play an important role at the phonemic level for lexical access. Our approach to Vietnamese expressive speech production consists using the concept of “rendez-vous” between linguistic levels and prosodic functions of utterance to combine the variation of tones and the global prosodic contours of expressive speech. However, as an under-resourced language, one main difficulty with Vietnamese speech processing is the lack of research and data, especially in the expressive speech domain.

In this presentation, we would like to present our work in 2nd years of PhD thesis. This work consists of construction and evaluation the first audio-visual expressive speech corpus for Vietnamese. A well-controlled recording methodology was designed to build a large representative audio-visual corpus for 16 attitudes, and one speaker. A perception experiment was carried out to evaluate a speaker’s perceived performances and to study the role and integration of the audio, visual, and audio-visual information in the listener’s perception of the speaker’s attitudes. The results reveal characteristics of Vietnamese prosodic attitudes and allow us to investigate such social affect in Vietnamese language.