Publications
Detail of publication
Citation
p. 283-290, Springer, Berlin-Heidelberg, Germany, 2010. : Listening-test-based annotation of communicative functions for expressive speech synthesis . Text, Speech and Dialogue, proceedings of the 13th International Conference TSD 2010, Lecture Notes in Artificial Intelligence, vol. 6231,
Download PDF
Additional information
Abstract
This paper is focused on the evaluation of listening test thatwas realized with a view to objectively annotate expressive speech recordingsand further develop a limited domain expressive speech synthesissystem. There are two main issues to face in this task. The first matterin issue to be taken into consideration is the fact that expressivity inspeech has to be defined in some way. The second problem is that perceptionof expressive speech is a subjective question. However, for thepurposes of expressive speech synthesis using unit selection algorithms,the expressive speech corpus has to be objectively and unambiguouslyannotated. At first, a classification of expressivity was determined makinguse of communicative functions. These are supposed to describe thetype of expressivity and/or speaker?s attitude. Further, to achieve objectivityat a significant level, a listening test with relatively high numberof listeners was realized. The listeners were asked to mark sentences inthe corpus using communicative functions. The aim of the test was toacquire a sufficient number of subjective annotations of the expressiverecordings so that we would be able to create ?objective? annotation.There are several methods to obtain objective evaluation from lots ofsubjective ones, two of them are presented.
Detail of publication
Title: | Listening-test-based annotation of communicative functions for expressive speech synthesis |
---|---|
Author: | Grůber, M. ; Matoušek, J. |
Language: | English |
Date of publication: | 6 Sep 2010 |
Year: | 2010 |
Type of publication: | Papers in journals |
Book title: | Text, Speech and Dialogue, proceedings of the 13th International Conference TSD 2010 |
Series: | Lecture Notes in Artificial Intelligence |
Číslo vydání: | 6231 |
Page: | 283 - 290 |
DOI: | 10.1007/978-3-642-15760-8_36 |
ISSN: | 0302-9743 |
Publisher: | Springer |
Address: | Berlin-Heidelberg, Germany |
Date: | 6 Sep 2010 - 10 Sep 2010 |
Keywords
expressive speech synthesis, listening test, communicativefunctions, inter-rater agreement measure
BibTeX
@INPROCEEDINGS{GruberM_2010_Listening-test-based, author = {Gr\r{u}ber, M. and Matou\v{s}ek, J.}, title = {Listening-test-based annotation of communicative functions for expressive speech synthesis}, year = {2010}, publisher = {Springer}, address = {Berlin-Heidelberg, Germany}, volume = {6231}, pages = {283-290}, booktitle = {Text, Speech and Dialogue, proceedings of the 13th International Conference TSD 2010}, series = {Lecture Notes in Artificial Intelligence}, ISSN = {0302-9743}, doi = {10.1007/978-3-642-15760-8_36}, url = {http://www.kky.zcu.cz/en/publications/GruberM_2010_Listening-test-based}, }