| 语音识别中的BP神经网络改进 |
| |
| 引用本文： | 张文静.语音识别中的BP神经网络改进[J].科学技术与工程,2007,7(18):4723-4725. |
| |
| 作者姓名： | 张文静 |
| |
| 作者单位： | 北京工业大学计算机学院,北京,100022;北京农业职业技术学院信息技术系,北京,102442 |
| |
| 摘 要： | 语音信号是受说话人、环境等因素影响的复杂函数，人工神经网络（ANN）模型进行语音识别具有自适应、自组织及联想等特征，适合用于语音识别。主要研究了传统BP神经网络在实际语音识别中的改进方案。利用改进后的方案实现语音识别、网络结构简化、运算速度加快、训练时间缩短。
|
| 关 键 词： | ANN BP算法 语音识别 规整网络 |
| 文章编号： | 1671-1819(2007)18-4723-03 |
| 修稿时间： | 2007-05-22 |
Improvement of BP Neural Network in Speech Recognition |
| |
| ZHANG Wen-jing.Improvement of BP Neural Network in Speech Recognition[J].Science Technology and Engineering,2007,7(18):4723-4725. |
| |
| Authors: | ZHANG Wen-jing |
| |
| Institution: | School of Computer, Beijing University of Technology, Beijing 100022, P. R. China; Technological Information Department Beijing Vocational College of Agriculture, Beijing 102442, P. R. China |
| |
| Abstract: | The speech signal is a complicated function influenced by factors such as the speaker,environment,etc.The artificial neural network(ANN)model carries on speech recognition with characteristics such as self-adaptation self-organization and self-imagination,it is suitable for speech recognition.The improvement scheme of traditional BP neural network in speech recognition is mainly studied.By making use of the scheme after improving,it can simplify network structure,accelerate operation,and shorten the training time. |
| |
| Keywords: | ANN back propagation algorithm speech recognition warping network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《科学技术与工程》浏览原始摘要信息 |
| 点击此处可从《科学技术与工程》下载免费的PDF全文 |
