Speech recognition using stereo vision neural networks with competition and cooperation

  • Authors:

  • Sung-III Kim

  • Affiliations:

  • Division of Electronic and Electrical Engineering, Kyungnam University, Masan City, Korea

  • Venue:
  • ISNN'05 Proceedings of the Second international conference on Advances in neural networks - Volume Part II
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper describes the speech recognition based on stereoscopic vision neural networks(SVNN) that has a dynamic process of self-organization that has been proved to be successful in recognizing a depth perception in stereoscopic vision. This study has shown that the process has also been useful in recognizing human speech. In the stereoscopic vision neural networks, the similarities are first obtained by comparing input vocal signals with standard models. They are then given to a dynamic process in which both competitive and cooperative processes are conducted among neighboring similarities. Finally, only one winner neuron is finally detected through the dynamic process. In a comparative study, the average phoneme recognition accuracies on the SVNN was 6.6 % higher than the existing recognizer based on Hidden Markov Models(HMM) with the structures of a single mixture and three states. From the results, therefore, it was noticed that the speech recognizer using SVNN outperformed the conventional recognizer in phoneme recognition under the same conditions.