Tone recognition for continuous Mandarin speech with limited training data using selected context-dependent hidden Markov models

被引：2

作者：

Wang, Hsin-Min ^{[1
]}

Lee, Lin-Shan ^{[1
]}

机构：

[1] Natl Taiwan Univ, Taipei, Taiwan

来源：

Journal of the Chinese Institute of Engineers, Transactions of the Chinese Institute of Engineers,Series A/Chung-kuo Kung Ch'eng Hsuch K'an | 1994年 / 17卷 / 06期

关键词：

Markov processes - Mathematical models - Selection - Speech;

D O I：

10.1080/02533839.1994.9677646

中图分类号：

学科分类号：

摘要：

Mandarin Chinese is a tonal language, in which every syllable is assigned a tone that has a lexical meaning. Therefore tone recognition is very important for Mandarin speech. This paper presents a method for continuous speech tone recognition. Context-dependent discrete hidden Markov models (HMM's) are used taking into account the tones of the syllables on both sides, and special efforts were made in selecting the minimum number of key context-dependent models considering the characteristics of the tones. The results indicate that a total of 23 context-dependent models have very good potential to describe the complicated tone behavior for all 175 possible tone concatenation conditions in continuous speech, such that the required training data can be reduced to a minimum and the recognition process can be simplified significantly. The best achievable recognition rate is 83.55%.

引用

页码：775 / 784

共 50 条

[41] Training hidden Markov models by hybrid simulated annealing for visual speech recognition
Jong-Seok Lee
Cheol Hoon Park
2006 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-6, PROCEEDINGS, 2006, : 198 - +
[42] Visual speech recognition using Active Shape Models and Hidden Markov Models
Luettin, J
Thacker, NA
Beet, SW
1996 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, CONFERENCE PROCEEDINGS, VOLS 1-6, 1996, : 817 - 820
[43] Landmine detection using ensemble discrete hidden Markov models with context dependent training methods
Hamdi, Anis
Missaoui, Oualid
Frigui, Hichem
Gader, Paul
DETECTION AND SENSING OF MINES, EXPLOSIVE OBJECTS, AND OBSCURED TARGETS XV, 2010, 7664
[44] Robust speech recognition using maximum likelihood neural networks and continuous density Hidden Markov Models
Yuk, DS
Che, CW
Flanagan, J
1997 IEEE WORKSHOP ON AUTOMATIC SPEECH RECOGNITION AND UNDERSTANDING, PROCEEDINGS, 1997, : 474 - 481
[45] TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS
Weber, P.
Houghton, S. M.
Champion, C. J.
Russell, M. J.
Jancovic, P.
2014 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2014,
[46] Visual speech recognition using motion features and hidden Markov models
Yau, Wai Chee
Kumar, Dinesh Kant
Weghorn, Hans
COMPUTER ANALYSIS OF IMAGES AND PATTERNS, PROCEEDINGS, 2007, 4673 : 832 - 839
[47] Telephone speech recognition using neural networks and hidden Markov models
Yuk, D
Flanagan, J
ICASSP '99: 1999 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, PROCEEDINGS VOLS I-VI, 1999, : 157 - 160
[48] Stereophonic speech recognition in noise using compensated hidden Markov models
Brookes, DM
Leung, MH
ELECTRONICS LETTERS, 1998, 34 (19) : 1827 - 1829
[49] Speech recognition using hidden Markov models based on segmental statistics
Toyohashi Univ of Technology, Toyohashi, Japan
Syst Comput Jpn, 7 (31-38):
[50] Telephone speech recognition using neural networks and hidden Markov models
Yuk, DongSuk
Flanagan, James
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, 1999, 1 : 157 - 160

← 1 2 3 4 5 →