Title :
A segment-based speech recognition system for isolated Mandarin syllables
Author :
Saga Chang ; Sin-Horng Chen
Author_Institution :
Dept. of Commun. Eng., Nat. Chiao Tung Univ., Hsinchu, Taiwan
Abstract :
A segment-based speech recognition scheme is proposed. The basic idea is to explicitly model the correlations between successive frames of an acoustic segment by using features representing the contours of spectral parameters. These segmental features are several lower-order coefficients of discrete orthonormal polynomial expansions. The performance of the proposed scheme was examined by simulations on multi-speaker speech recognition for all 408 highly confusing first-tone Mandarin syllables. A recognition rate of 77.4% was achieved for the case, using five 6-segment reference templates per syllable. This is 13.0% and 6.6% higher than those obtained by a conventional dynamic time warping (DTW) method and a conventional hidden Markov model (CHMM) method, respectively.<>
Keywords :
speech recognition; Chinese; acoustic segment; discrete orthonormal polynomial expansions; dynamic time warping; first-tone Mandarin syllables; hidden Markov model; isolated Mandarin syllables; lower-order coefficients; multi-speaker speech recognition; performance; recognition rate; reference templates; segment-based speech recognition system; simulations; spectral parameter contours; successive frame correlations; Acoustic measurements; Acoustical engineering; Data mining; Filter bank; Hidden Markov models; Linear predictive coding; Phase measurement; Signal analysis; Speech analysis; Speech recognition;
Conference_Titel :
TENCON '93. Proceedings. Computer, Communication, Control and Power Engineering.1993 IEEE Region 10 Conference on
Conference_Location :
Beijing, China
Print_ISBN :
0-7803-1233-3
DOI :
10.1109/TENCON.1993.327986
