Title :
Combining multi-section Bayesian template with level-building algorithm for robust connected Mandarin digit recognition
Author :
Shyu, Ruey-Ching ; Wang, Jhing-Fa ; Huang, Chaug-Ching ; Wu, Chung-Hsien ; Shyuu, Jyh-Shing ; Lee, Jau-Yien
Author_Institution :
Nat. Cheng Kung Univ., Tainan, Taiwan
Abstract :
A robust connected Mandarin digit recognition system based on multi-section Bayesian templates and the level-building algorithm is presented. In the multi-section Bayesian template, both coarticulation effects and the characteristics of the digit itself can be properly exhibited. In particular, each section is organized into a Bayesian template which is characterized by a powerful statistical framework. In the recognition phase, several factors such as finding a more desirable lower and upper bound of input speech length, word duration checking and post-processing are carefully considered to make the proposed system more robust. The proposed system is tested using a multi-speaker database (25 males, 25 females) and a string accuracy of 95.4% is achieved.
Keywords :
Bayes methods; speech analysis and processing; speech recognition; coarticulation effects; connected Mandarin digit recognition system; input speech length; level-building algorithm; multi-section Bayesian templates; multi-speaker database; speech recognition; statistical framework; string accuracy; word duration; Bayesian methods; Databases; Iterative algorithms; Maximum likelihood estimation; Natural languages; Pattern recognition; Robustness; Speech recognition; Testing; Upper bound;
Conference_Titel :
VLSI Technology, Systems, and Applications, 1993. Proceedings of Technical Papers. 1993 International Symposium on
Conference_Location :
Taipei, Taiwan
Print_ISBN :
0-7803-0978-2
DOI :
10.1109/VTSA.1993.263603