doi: 10.4304/jsw.7.4.779-785
Automatic Feature Template Generation for Prosodic Phrasing
2Department of Applied Mathematics, Hunan University of Finance and Economics, Changsha, China
Abstract—Prosodic phrase prediction is important for both the naturalness and intelligibility of Text-to-Speech (TTS) systems. To automatically generate feature templates of prosodic phrasing models, this paper proposes a hybrid approach which converts the rules generated by classification and regression tree (CART) into templates of transformation-based learning (TBL), and designs a hierarchical clustering based feature combination algorithm for maximum entropy (ME) model. While minimizing human supervision, TBL templates automatically generated by CART can provide good alternatives or beneficial supplement to manually summarized templates, and ME templates automatically generated by the proposed feature combination algorithm not only make an improvement of 3.1% on F-measure over manual templates, but also reduce the size of ME model by up to 79.0%.
Index Terms—prosodic phrase prediction, feature template generation, keyword selection, classification and regression tree (CART), transformation-based learning (TBL), maximum entropy (ME).
Cite: Fangzhou Liu and You Zhou, "Automatic Feature Template Generation for Prosodic Phrasing," Journal of Software vol. 7, no. 4, pp. 779-785, 2012.
General Information
ISSN: 1796-217X (Online)
Abbreviated Title: J. Softw.
Frequency: Quarterly
APC: 500USD
DOI: 10.17706/JSW
Editor-in-Chief: Prof. Antanas Verikas
Executive Editor: Ms. Cecilia Xie
Abstracting/ Indexing: DBLP, EBSCO,
CNKI, Google Scholar, ProQuest,
INSPEC(IET), ULRICH's Periodicals
Directory, WorldCat, etcE-mail: jsweditorialoffice@gmail.com
-
Oct 22, 2024 News!
Vol 19, No 3 has been published with online version [Click]
-
Jan 04, 2024 News!
JSW will adopt Article-by-Article Work Flow
-
Apr 01, 2024 News!
Vol 14, No 4- Vol 14, No 12 has been indexed by IET-(Inspec) [Click]
-
Apr 01, 2024 News!
Papers published in JSW Vol 18, No 1- Vol 18, No 6 have been indexed by DBLP [Click]
-
Jun 12, 2024 News!
Vol 19, No 2 has been published with online version [Click]