Parallel sequential running neural network and its application to automatic speech recognition

Author

Zeng, Huaiyu ; Yu, Tiecheng

Author_Institution

Inst. of Acoust., Acad. Sinica, Beijing, China

Volume

1

fYear

1992

fDate

23-26 Mar 1992

Firstpage

429

Abstract

A novel parallel sequential running neural network (PSRNN) is developed. It consists of subnets of the same construction. The subnet was trained by different tokens sequentially. The neural network makes recognition by subnets in the order of training. PSRNN performs better than multilayer perceptron (MLP). It can learn adaptively and expand easily. The authors applied PSRNN to the work of speaker-independent isolated word recognition. The system was trained by 45 persons to recognize ten Chinese digits. Performance was 97% when tested by another 10 persons

Keywords

learning (artificial intelligence); neural nets; speech recognition; Chinese digits; adaptive learning; automatic speech recognition; parallel sequential running neural network; sequential training; speaker-independent isolated word recognition; subnets; tokens; Acoustic applications; Artificial neural networks; Automatic speech recognition; Biological neural networks; Error analysis; Neural networks; Neurons; Performance evaluation; Speech recognition; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on

Conference_Location

San Francisco, CA

ISSN

1520-6149

Print_ISBN

0-7803-0532-9

Type

conf

DOI

10.1109/ICASSP.1992.225880

Filename

225880