Continuous speech research based on two-weight neural network

Authors:
Wenming Cao;Xiaoxia Pan;Shoujue Wang
Affiliations:
Information College, Zhejiang University of Technology, Hangzhou, Zhejiang, China;Information College, Zhejiang University of Technology, Hangzhou, Zhejiang, China;Institute of Semiconductors, Chinese Academy of Sciences, Beijing, China
Venue:
ISNN'05 Proceedings of the Second international conference on Advances in neural networks - Volume Part II
Year:
2005

Citing 1
Cited 4

The application of DBF neural networks for object recognition

Information Sciences—Informatics and Computer Science: An International Journal

Channel equalization based on two weights neural network

CIS'05 Proceedings of the 2005 international conference on Computational Intelligence and Security - Volume Part II
Continuous speech research based on hypersausage neuron

CIS'05 Proceedings of the 2005 international conference on Computational Intelligence and Security - Volume Part II
High-Dimensional space geometrical informatics and its applications to image restoration

ISNN'06 Proceedings of the Third international conference on Advnaces in Neural Networks - Volume Part II
A novel image restoration algorithm based on high-dimensional space geometry

ICCSA'06 Proceedings of the 2006 international conference on Computational Science and Its Applications - Volume Part V

Quantified Score

Hi-index	0.00

Visualization

Abstract

Two-weight neural network (TWNN) is described in this paper. A new dynamic searching algorithm based on Two-weight neural network is presented. And then it is applied to recognize the Continuous Speech of Speaker-Independent. The recognition results can be searched dynamically without endpoint detecting and segmenting. Different feature-space covers are constructed according to different classes of syllables. Compared with the conventional HMM-based method, The trend of recognition results shows that the difference of recognition rates between these two methods decreases as the number of training increases, but the recognition rate of Two-weight neural network is always higher than that of HMM-based. And both of these recognition rates will reach 100% if there are enough training samples.