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

  • Authors:

  • Huaiyu Zeng;Tiecheng Yu

  • Affiliations:

  • Institute of Acoustics, Academia Sinica, Beijing, P.R.C.;Institute of Acoustics, Academia Sinica, Beijing, P.R.C.

  • Venue:
  • ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1
  • Year:
  • 1992

Quantified Score

Hi-index 0.00

Visualization

Abstract

A novel parallel sequential running neural network (PSRNN) is developed. It consists of subnets of same construction. The subnet was trained by different tokens sequentially. The neural network make recognition by subnets in the order of training. PSRNN performs better than MLP. It can learn adaptively and expand easily. We 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.