DocumentCode :
2496518
Title :
Mandarin stress detection using hierarchical model based boosting classification and regression tree
Author :
Ni, Chong-Jia ; Liu, Wen-Ju ; Xu, Bo
Author_Institution :
Nat. Lab. of Pattern Recognition, Chinese Acad. of Sci., Beijing, China
fYear :
2010
fDate :
18-23 July 2010
Firstpage :
1
Lastpage :
5
Abstract :
Automatic stress detection is important for both speech understanding and natural speech synthesis. In this paper, we develop hierarchical model based boosting classification and regression tree (CART) to detect Mandarin stress by using acoustic evidence and text information. When comparing with previous proposed method at the same training and test sets, there are 2.52% and 1.09% absolute accuracy rate improvements respectively. We also analyze the differences between Mandarin stress detection and English pitch accent prediction, and prove some linguistic conclusions based on the large corpus in a different way.
Keywords :
linguistics; natural language processing; pattern classification; regression analysis; speech processing; speech synthesis; trees (mathematics); English pitch accent prediction; Mandarin stress detection; acoustic evidence; hierarchical model based boosting classification; natural speech synthesis; regression tree; speech understanding; text information; Predictive models; Presses; Support vector machines;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks (IJCNN), The 2010 International Joint Conference on
Conference_Location :
Barcelona
ISSN :
1098-7576
Print_ISBN :
978-1-4244-6916-1
Type :
conf
DOI :
10.1109/IJCNN.2010.5596862
Filename :
5596862
Link To Document :
بازگشت