首页 | 本学科首页   官方微博 | 高级检索  
     

BP神经网络应用于孤立词语发音识别的研究
引用本文:姜占才,孙燕.BP神经网络应用于孤立词语发音识别的研究[J].计算机应用与软件,2008,25(10).
作者姓名:姜占才  孙燕
作者单位:1. 青海师范大学物理系,青海,西宁,810008
2. 青海民族学院计算机科学与技术系,青海,西宁,810007
摘    要:介绍了BP神经网络的学习规则和用于语音识别的基本原理,建立了一个用于常用孤立词语音识别的BP神经网络,选择声道反射系数为语音识别的特征值,建立了网络的训练样本集,对网络进行了训练;用MATLAB进行了识别仿真,表明能较好地实现孤立词语音识别.

关 键 词:孤立词  训练  反射系数  BP神经网络  语音识别

ON APPLYING BP NEURAL NETWORK TO ISOLATED WORDS SPEECH RECOGNITION
Jiang Zhancai,Sun Yan.ON APPLYING BP NEURAL NETWORK TO ISOLATED WORDS SPEECH RECOGNITION[J].Computer Applications and Software,2008,25(10).
Authors:Jiang Zhancai  Sun Yan
Affiliation:Jiang Zhancai1 Sun Yan21(Department of Physics,Qinghai Normal University,Xining 810008,Qinghai,China)2(Department of Computer Science , Technology,Qinghai University for Nationalities,Xining 810007,China)
Abstract:The learning principle of BP Artificial Neural Network(ANN) and the basic theory on speech recognition is introduced in this article.A BP Artificial Neural Network for common isolated word recognition is presented.Selecting reflection coefficients of vocal track as feature value of the speech recognition to construct the training sample set of BP Artificial Neural Network,the ANN is trained.The recognition results obtained from MATLAB simulation show that the speech recognition of isolated words achieves we...
Keywords:Isolated words Train Reflection coefficients BP artificial neural network Speech recognition  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号