Skip to content

Publications

Detail of publication

Citation

Hanzlíček, Z. and Grůber, M. : Initial Experiments on Automatic Correction of Prosodic Annotation of Large Speech Corpora . Text, Speech, and Dialogue, Proceedings of the 17th International Conference TSD 2014, Lecture Notes in Artificial Intelligence, vol. 8655, p. 481-488, Springer International Publishing, Berlin-Heidelberg, Germany, 2014.

Download PDF

PDF

Abstract

Most modern speech synthesis systems utilize large speech corpora to learn new voices. These speech corpora usually contain several hours of speech spoken by talented speakers who are able to record such an amount of speech data in a sufficient quality. An appropriate phonetic and prosodic annotation of the recorded utterances is necessary for a high quality of synthesized speech. For many languages, the pitch shape within the last prosodic word of a phrase is characteristic for particular types of sentences and phrase structure of compound/complex sentences. However in the real data, this formal convention can be breached and a different pitch shape than expected can be present. This can be a source of prosody inconsistency in synthesized speech. This article presents some experiments on automatic detection of prosodic mismatch in recorded utterances. A simple classifier based on GMM was proposed for this task. Experiments were performed on 5 large speech corpora. The classification results were successfully verified by listening tests.

Detail of publication

Title: Initial Experiments on Automatic Correction of Prosodic Annotation of Large Speech Corpora
Author: Hanzlíček, Z. ; Grůber, M.
Language: English
Date of publication: 1 Sep 2014
Year: 2014
Type of publication: Papers in journals
Book title: Text, Speech, and Dialogue, Proceedings of the 17th International Conference TSD 2014
Series: Lecture Notes in Artificial Intelligence
Číslo vydání: 8655
Page: 481 - 488
DOI: 10.1007/978-3-319-10816-2_58
ISBN: 978-3-319-10815-5
ISSN: 0302-9743
Publisher: Springer International Publishing
Address: Berlin-Heidelberg, Germany
Date: 8 Sep 2014 - 12 Sep 2014
/ 2017-03-22 08:53:38 /

Keywords

speech corpora, prosodic annotation, prosodeme

BibTeX

@INCOLLECTION{HanzlicekZ_2014_InitialExperiments,
 author = {Hanzl\'{i}\v{c}ek, Z. and Gr\r{u}ber, M.},
 title = {Initial Experiments on Automatic Correction of Prosodic Annotation of Large Speech Corpora },
 year = {2014},
 publisher = {Springer International Publishing},
 address = {Berlin-Heidelberg, Germany},
 volume = {8655},
 pages = {481-488},
 booktitle = {Text, Speech, and Dialogue, Proceedings of the 17th International Conference TSD 2014},
 series = {Lecture Notes in Artificial Intelligence},
 ISBN = {978-3-319-10815-5},
 ISSN = {0302-9743},
 doi = {10.1007/978-3-319-10816-2_58},
 url = {http://www.kky.zcu.cz/en/publications/HanzlicekZ_2014_InitialExperiments},
}