DocumentCode :
3424359
Title :
Acoustic modeling with contextual additive structure for HMM-based speech recognition
Author :
Nankaku, Yoshihiko ; Nakamura, Kazuhiro ; Zen, Heiga ; Tokuda, Keiichi
Author_Institution :
Dept. of Comput. Sci. & Eng., Nagoya Inst. of Technol., Nagoya
fYear :
2008
fDate :
March 31 2008-April 4 2008
Firstpage :
4469
Lastpage :
4472
Abstract :
This paper proposes an acoustic modeling technique based on an additive structure of context dependencies for HMM-based speech recognition. Typical context dependent models, e.g., triphone HMMs, have direct dependencies of phonetic contexts, i.e., if a phonetic context is given, the Gaussian distribution is specified immediately. This paper assumes a more complex structure, an additive structure of acoustic feature components which have different context dependencies. Since the output probability distribution is composed of additive component distributions, a number of different distributions can be efficiently represented by a combination of fewer distributions. To automatically extract additive components, this paper presents a context clustering algorithm for the additive structure model in which multiple decision trees are constructed simultaneously. Experimental results show that the proposed technique improves phoneme recognition accuracy with fewer number of distributions than the conventional triphone HMMs.
Keywords :
Gaussian distribution; acoustic signal processing; decision trees; hidden Markov models; speech processing; speech recognition; Gaussian distribution; HMM-based speech recognition; acoustic feature components; acoustic modeling; additive component distributions; complex structure; context clustering algorithm; contextual additive structure; multiple decision trees; phoneme recognition accuracy; phonetic contexts; probability distribution; Additives; Clustering algorithms; Computer science; Context modeling; Decision trees; Hidden Markov models; Linear regression; Probability distribution; Speech recognition; Training data; Additive structure; Context clustering; Decision trees; Distribution convolution; Hidden Markov models;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech and Signal Processing, 2008. ICASSP 2008. IEEE International Conference on
Conference_Location :
Las Vegas, NV
ISSN :
1520-6149
Print_ISBN :
978-1-4244-1483-3
Electronic_ISBN :
1520-6149
Type :
conf
DOI :
10.1109/ICASSP.2008.4518648
Filename :
4518648
Link To Document :
بازگشت