Title :
A new neural network oriented speech recognition
Author :
Xiaoming, Wang ; Baoyu, Zheng
Author_Institution :
Nanjing Univ. of Posts & Telecommun., China
Abstract :
A new neural network is proposed. The training and recognition algorithms are described. In order to improve the ability of modeling the temporal signal, a preprocessing approach named nonlinear segmentation and representing-frame feature extraction is employed. A Chinese speech recognition system based on the new neural network is established and has achieved high performance
Keywords :
feature extraction; learning (artificial intelligence); neural nets; signal representation; speech recognition; Chinese speech recognition system; high performance; neural network oriented speech recognition; nonlinear segmentation; preprocessing approach; recognition algorithm; representing-frame feature extraction; temporal signal modeling; training algorithm; Artificial neural networks; Credit cards; Feature extraction; Hidden Markov models; Information processing; Neural networks; Pattern matching; Signal processing; Speech processing; Speech recognition;
Conference_Titel :
Communication Technology Proceedings, 1998. ICCT '98. 1998 International Conference on
Conference_Location :
Beijing
Print_ISBN :
7-80090-827-5
DOI :
10.1109/ICCT.1998.741139
