Performance comparison of several speech recognition methods

Author

Guan, Cuntai ; Zhu, Ce ; Chen, Yongbin ; He, Zhenya

Author_Institution

Dept. of Radio Eng., Southeast Univ., Nanjing, China

fYear

1994

fDate

13-16 Apr 1994

Firstpage

710

Abstract

The paper compares several hybrid speech recognition methods an a highly confusable recognition vocabulary set. To characterize spectral variability of speech in speech recognition, Dynamic spectral warping (DSW) method is combined with DTW, HMM, and LVQ2 models. With these combinations, the discriminative abilities of these models are improved. Recognition accuracy is improved when these modified methods are tested on a speaker-dependent Chinese consonant recognition task. A TDNN model is also tested on the same database for comparison

Keywords

hidden Markov models; learning (artificial intelligence); natural languages; spectral analysis; speech recognition; vector quantisation; HMM; LVQ2 model; TDNN model; confusable recognition vocabulary set; dynamic spectral warping; hidden Markov model; learning vector quantization; performance comparison; speaker-dependent Chinese consonant recognition task; spectral variability; speech recognition methods; Databases; Dynamic programming; Helium; Hidden Markov models; Image processing; Multi-layer neural network; Neural networks; Speech processing; Speech recognition; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Speech, Image Processing and Neural Networks, 1994. Proceedings, ISSIPNN '94., 1994 International Symposium on

Print_ISBN

0-7803-1865-X

Type

conf

DOI

10.1109/SIPNN.1994.344813

Filename

344813