DocumentCode :
1857830
Title :
Acoustic-syntactic maximum entropy model for automatic prosody labeling
Author :
Rangarajan, V. ; Narayanan, S. ; Bangalore, S.
Author_Institution :
Viterbi Sch. of Electr. Eng., Southern California Univ., Los Angeles, CA
fYear :
2006
fDate :
10-13 Dec. 2006
Firstpage :
74
Lastpage :
77
Abstract :
In this paper we describe an automatic prosody labeling framework that exploits both language and speech information intended for the purpose of incorporating prosody within a speech-to-speech translation framework. We propose a maximum entropy syntactic- prosodic model that achieves an accuracy of 85.22% and 91.54% for pitch accent and boundary tone labeling on the Boston University Radio News corpus. We model the acoustic-prosodic stream with two different models, one a maximum entropy model and the other a traditional HMM. We finally couple the syntactic-prosodic and acoustic-prosodic components to achieve a pitch accent and boundary tone classification accuracy of 86.01% and 93.09% respectively.
Keywords :
language translation; maximum entropy methods; natural languages; speech processing; acoustic-prosodic component; acoustic-syntactic maximum entropy model; automatic prosody labeling; language information; speech information; speech-to-speech translation framework; syntactic-prosodic component; Entropy; Equations; Hidden Markov models; Labeling; Loudspeakers; Natural languages; Signal synthesis; Speech analysis; Speech recognition; Speech synthesis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Spoken Language Technology Workshop, 2006. IEEE
Conference_Location :
Palm Beach
Print_ISBN :
1-4244-0872-5
Type :
conf
DOI :
10.1109/SLT.2006.326820
Filename :
4123365
Link To Document :
بازگشت