基于语音识别和文本分词算法的指令交互系统的设计和实现.doc

文档介绍：
基于语音识别和文本分词算法的指令交互
系统的设计和实现
张文杰,张洪刚,闫月娥**
5
10
15
20
25
30
35
40
(北京邮电大学信息与通信工程学院)
摘要:由于军事作战的多变性和复杂性,在用计算机进行军事指挥操作的过程中,指挥官迫
切希望能使用一种更加方便、快捷的方式。在此基础上,本文提出了一种采用语音技术和分
词技术相结合的指令交互系统,该系统采用关键词识别率较高的 Speech SDK 进行语音识别,
同时在结合军事指挥系统操作习惯的基础上,采取了一种基于改进的整词二分的词典机制的
正向最大匹配算法。该方法设计的军事指挥系统具有很高的实用性和灵活性。
关键词:语音识别,Speech SDK,正向最大匹配,哈希散列表
中图分类号:TP391
The Design and Implementation of mands
interactive system based on Speech Recognitoon and Text
Segmentation
ZHANG Wenjie1, ZHANG Honggang2, YAN Yuee1
(1. School of Information munication Engineering,BeijingUniversity Posts and
munications;
2. School of Information munication Engineering,Beijing University Posts and
munications)
Abstract: As variability plexity of military operations,during the operation of military
command manders are eager to be able to use a more convenient and efficient
way. On this basis, this paper presents a mands interactive system bases on speech
recognition and chinese word segmentation, This system uses Speech SDK for speech recogniton
which has a higher keyword recognition rate, at the same time, on the basis of customary of the
mand system operating,it adopts the forward maximum matching algorithm based on
an improved Dictionary Mechanism. This method is designed for the mand system
with high availability and flexibility.
Key words: Speech Recognition, Speech SDK, Forward maximum mat