Title: | Identification and automatic detection of parasitic speech sounds |
Other Titles: | Identifikace a automatická detekce parazitních řečových zvuků |
Authors: | Matoušek, Jindřich Skarnitzl, Radek Machač, Pavel Trmal, Jan |
Citation: | MATOUŠK, Jindřich; SKARNITZL, Radek; MACHAČ, Pavel; TRMAL, Jan. Identification and automatic detection of parasitic speech sounds. In: Proceedings of ICSPL 2009: 10th Annual Conference of the International Speech Communication Association 2009, 6-10 September 2009, Brighton, UK. [Baixas]: ISCA, 2009, p. 876-879. ISSN 1990-9772. |
Issue Date: | 2009 |
Publisher: | ISCA |
Document type: | článek article |
URI: | http://www.kky.zcu.cz/cs/publications/MatousekJ_2009_Identificationand http://hdl.handle.net/11025/17018 |
ISSN: | 1990-9772 |
Keywords: | parazitní zvuk řeči;lingvistická přirozenost;syntéza řeči;výbě›r jednotky;HMM;BVM |
Keywords in different language: | parasitic speech sound;linguistic naturalness;speech synthesis;unit selection;HMM;BVM |
Abstract in different language: | This paper presents initial experiments with the identification and automatic detection of parasitic sounds in speech signals. The main goal of this study is to identify such sounds in the source recordings for unit-selection-based speech synthesis systems and thus to avoid their unintended usage in synthesised speech. The first part of the paper describes the phonetic analysis and identification of parasitic phenomena in recordings of two Czech speakers. In the second part, experiments with the automatic detection of parasitic sounds using HMM-based and BVM classifiers are presented. The results are encouraging, especially those for glottalization phenomena. |
Rights: | © Jindřich Matoušek - Radek Skarnitzl - Pavel Machač - Jan Trmal |
Appears in Collections: | Články / Articles (KKY) Články / Articles (KIV) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
MatousekJ_2009_Identificationand.pdf | Plný text | 398,8 kB | Adobe PDF | View/Open |
Please use this identifier to cite or link to this item:
http://hdl.handle.net/11025/17018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.