Automatic feature template generation for maximum entropy based intonational phrase break prediction

You Zhou

doi:10.1117/12.2010767

14 March 2013 Automatic feature template generation for maximum entropy based intonational phrase break prediction

You Zhou

Proceedings Volume 8768, International Conference on Graphic and Image Processing (ICGIP 2012); 87681H (2013) https://doi.org/10.1117/12.2010767
Event: 2012 International Conference on Graphic and Image Processing, 2012, Singapore, Singapore

Abstract

The prediction of intonational phrase (IP) breaks is important for both the naturalness and intelligibility of Text-to- Speech (TTS) systems. In this paper, we propose a maximum entropy (ME) model to predict IP breaks from unrestricted text, and evaluate various keyword selection approaches in different domains. Furthermore, we design a hierarchical clustering algorithm for automatic generation of feature templates, which minimizes the need for human supervision during ME model training. Results of comparative experiments show that, for the task of IP break prediction, ME model obviously outperforms classification and regression tree (CART), log-likelihood ratio is the best scoring measure of keyword selection, compared with manual templates, templates automatically generated by our approach greatly improves the F-score of ME based IP break prediction, and significantly reduces the size of ME model.

Citation Download Citation

You Zhou "Automatic feature template generation for maximum entropy based intonational phrase break prediction", Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87681H (14 March 2013); https://doi.org/10.1117/12.2010767

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available