Skip to content

Publications

Detail of publication

Citation

Matoušek, J and Tihelka, D. : Anomaly-Based Annotation Errors Detection in TTS Corpora . Proceedings of the 16th Annual Conference of the International Speech Communication Association (Interspeech 2015), p. 314-318, Dresden, Germany, 2015.

Additional information


ISCA archive

Abstract

n this paper we adopt several anomaly detection methods to detect annotation errors in single-speaker read-speech corpora used for text-to-speech (TTS) synthesis. Correctly annotated words are considered as normal examples on which the detec- tion methods are trained. Misannotated words are then taken as anomalous examples which do not conform to normal patterns of the trained detection models. Word-level feature sets including basic features derived from forced alignment, and various acoustic, spectral, phonetic, and positional features were examined. Dimensionality reduction techniques were also applied to reduce the number of features. The first results with F1 score being almost 89% show that anomaly detection could help in detecting annotation errors in read-speech corpora for TTS synthesis.

Detail of publication

Title: Anomaly-Based Annotation Errors Detection in TTS Corpora
Author: Matoušek, J ; Tihelka, D.
Language: English
Date of publication: 6 Sep 2015
Year: 2015
Type of publication: Papers in proceedings of reviewed conferences
Book title: Proceedings of the 16th Annual Conference of the International Speech Communication Association (Interspeech 2015)
Page: 314 - 318
ISSN: 2308-457X
Address: Dresden, Germany
Date: 6 Sep 2015 - 10 Sep 2015
/ 2016-01-13 17:33:14 /

Keywords

annotation error detection, anomaly detection, read speech corpora, speech synthesis

BibTeX

@INPROCEEDINGS{MatousekJ_2015_Anomaly-Based,
 author = {Matou\v{s}ek, J and Tihelka, D.},
 title = {Anomaly-Based Annotation Errors Detection in TTS Corpora},
 year = {2015},
 address = {Dresden, Germany},
 pages = {314-318},
 booktitle = {Proceedings of the 16th Annual Conference of the International Speech Communication Association (Interspeech 2015)},
 ISSN = {2308-457X },
 url = {http://www.kky.zcu.cz/en/publications/MatousekJ_2015_Anomaly-Based},
}