Napredna pretraga

Pregled bibliografske jedinice broj: 554120

Automatic Intonation Event Detection Using Tilt Model for Croatian Speech Synthesis


Načinović, Lucia; Pobar, Miran; Martinčić-Ipšić, Sanda; Ipšić, Ivo
Automatic Intonation Event Detection Using Tilt Model for Croatian Speech Synthesis // INFuture2011, The Future of Information Sciences, Information Sciences and e-Society / Billenness, Clive ; Hemera, Annette ; Mateljan, Vladimir ; Banek Zorica, Mihaela ; Stančić, Hrvoje ; Seljan, Sanja (ur.).
Zagreb: Department of Information Sciences, Faculty of Humanities and Social Sciences, University of Zagreb, 2011. str. 383-391 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Automatic Intonation Event Detection Using Tilt Model for Croatian Speech Synthesis

Autori
Načinović, Lucia ; Pobar, Miran ; Martinčić-Ipšić, Sanda ; Ipšić, Ivo

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
INFuture2011, The Future of Information Sciences, Information Sciences and e-Society / Billenness, Clive ; Hemera, Annette ; Mateljan, Vladimir ; Banek Zorica, Mihaela ; Stančić, Hrvoje ; Seljan, Sanja - Zagreb : Department of Information Sciences, Faculty of Humanities and Social Sciences, University of Zagreb, 2011, 383-391

ISBN
978-953-175-408-8

Skup
INFuture2011

Mjesto i datum
Zagreb, Hrvatska, 09-11.11.2011

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Prosody in TTS; intonation model; Tilt

Sažetak
Text-to-speech systems convert text into speech. Synthesized speech without prosody sounds unnatural and monotonous. In order to sound natural, prosodic elements have to be implemented. The generation of prosodic elements directly from text is a rather demanding task. Our final goals are building a complete prosodic model for Croatian and implementing it into our TTS system. In this work, we present one of the steps in implementation of prosody into TTSs – detection of intonation events using Tilt intonation model. We propose a training procedure which is composed of several subtasks. First, we hand-labelled a set of utterances and within each of them, marked four types of prosodic events. Then we trained HMMs and used them to mark prosodic events on a larger set of utterances. Finally, we assigned parameters to each of the event and generated f0 contours from the parameters .We evaluated the obtained f0 contours.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti



POVEZANOST RADA


Projekt / tema
318-0361935-0852 - Govorne tehnologije (Ivo Ipšić, )

Ustanove
Sveučilište u Rijeci - Odjel za informatiku