DocumentCode :
431243
Title :
DBN-Based Multi-stream Models for Mandarin Toneme Recognition
Author :
Lei, Xin ; Ji, Gang ; Ng, Tim ; Bilmes, Jeff ; Ostendorf, Mari
Author_Institution :
Dept. of Electr. Eng., Univ. of Washington, Seattle, WA, USA
Volume :
1
fYear :
2005
fDate :
March 18-23, 2005
Firstpage :
349
Lastpage :
352
Keywords :
belief networks; cepstral analysis; feature extraction; hidden Markov models; speech processing; speech recognition; DBN-based multi-stream models; Graphical Model Toolkit; HMM; MFCC; Mandarin Chinese; Mandarin toneme recognition; asynchronous multi-stream systems; base phone; concatenated features; dynamic Bayesian networks; pitch features; recognition systems; stream-level model parameter tying; synchronous multi-stream systems; tonal phone; tones; Bayesian methods; Cepstral analysis; Concatenated codes; Decoding; Graphical models; Hidden Markov models; Mel frequency cepstral coefficient; Natural languages; Random variables; Speech recognition;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech, and Signal Processing, 2005. Proceedings. (ICASSP '05). IEEE International Conference on
ISSN :
1520-6149
Print_ISBN :
0-7803-8874-7
Type :
conf
DOI :
10.1109/ICASSP.2005.1415122
Filename :
1415122
Link To Document :
بازگشت