An improvement of initial-final based Mandarin continuous speech recognition
碩士 === 國立交通大學 === 電信研究所 === 83 === In this thesis, several techniques to improve the initial-final based HMM method for continuous Mandarin speech recognition are proposed. The baseline system uses 100 right-context-dependent initial HMM models and 39 con...
Main Authors: | S. M. Chiang, 蔣松茂 |
---|---|
Other Authors: | S. H. Chen |
Format: | Others |
Language: | zh-TW |
Published: |
1995
|
Online Access: | http://ndltd.ncl.edu.tw/handle/61597883163258818852 |
Similar Items
-
Initial-Final Modeling and Recognition of Mandarin Speech
by: Chou, Chia-Shyan, et al.
Published: (1996) -
Continuous Mandarin Syllable Recognition for Telephone Speech
by: Chiang, Sheng-lun, et al.
Published: (1997) -
INITIAL/FINAL Confusing Sets in Mandarin Speech Recognition
by: Yen-Chi Lee, et al.
Published: (1999) -
An Improvement on the HMM-based Continuous Mandarin Speech Recognition Method
by: Chu-Yang Lu, et al.
Published: (2002) -
Improved discriminative training for Mandarin continuous speech recognition
by: Shih-Hung Liu, et al.
Published: (2007)