首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于Windows的离线关键词语音识别系统设计与实现
引用本文:孙林超,秦会斌,崔佳冬.基于Windows的离线关键词语音识别系统设计与实现[J].教育技术导刊,2019,18(9):116-120.
作者姓名:孙林超  秦会斌  崔佳冬
作者单位:杭州电子科技大学 新型电子器件与应用研究所,浙江 杭州 310018
摘    要:由于传统人机交互大多使用键盘、鼠标等交互方式,速度较慢,因此语音识别开始受到越来越多人的青睐。但语音识别也存在如扩展性太差、可复制性不好造成单个产品价格过高、过于依赖外部条件导致对自身使用有所限制等问题。设计并实现一种基于本地的语音识别系统,通过构建抽象语法树,实现语音控制操作。实验结果表明,该系统的离线识别准确率可达70%以上,可以在局域网内实现语音操作。

关 键 词:语音识别  离线识别  XML文件  语法树    
收稿时间:2019-07-16

Design and Implementation of Offline Keyword Recognition Based on Windows
SUN Lin-chao,QIN Hui-bin,CUI Jia-dong.Design and Implementation of Offline Keyword Recognition Based on Windows[J].Introduction of Educational Technology,2019,18(9):116-120.
Authors:SUN Lin-chao  QIN Hui-bin  CUI Jia-dong
Institution:Institute of Electron Devices & Application, Hangzhou Dianzi University,Hangzhou 310018, China
Abstract:In view of the shortcomings of the keyboard and mouse interaction methods used in the traditional human-computer interaction mode, such as slow speed and easy travel, voice recognition is increasingly favored by everyone. Nowadays, there are different problems in the speech recognition, such as extremely poor expansion, poor reproducibility of individual products, and relatively limited dependence on external conditions. We have designed and implemented a local speech recognition system in which an abstract grammar tree is constructed to realize operation controlled by speech. The experimental results show that the accuracy of the design is over 70%, and the effective speech recognition is realized. By setting off-line keywords with good extensibility and reproducibility, we can achieve the independence of speech recognition and the desired voice operation in the LAN.
Keywords:speech recognition  offline recognition  XML file  grammar tree  
点击此处可从《教育技术导刊》浏览原始摘要信息
点击此处可从《教育技术导刊》下载免费的PDF全文
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