语音特征和情感特征的翻译系统与实现  

Translation system based on speech feature and emotion feature and its implementation

作　　者：曹春香[1] CAO Chunxiang(Weifang Medical University,Weifang 261053,China)

机构地区：[1]潍坊医学院,山东潍坊261053

出　　处：《现代电子技术》2018年第13期123-127,共5页Modern Electronics Technique

摘　　要：传统在线机器翻译系统存在翻译效率低、翻译效果差等问题,因此设计了基于语音特征和情感特征的翻译系统。系统通过语音情感特征提取模块将识别出的语音情感特征的具体特征提取出来,在模型库中进行特征分类,提取出的特征信息在翻译模块中进行翻译处理。翻译模块由用户监视器、内核监视器、CPU模拟器、指令翻译器、外设模拟器以及翻译处理器构成,实现翻译目标的准确翻译。通过语音特征识别模块的实现完成系统语音和情感特征的准确识别和检测,通过翻译处理器得到准确的翻译结果。实验结果表明该系统翻译准确率高,翻译效果好,系统响应时间短。The traditional online machine translation system has low translation efficiency and poor translation effect,therefore a translation system based on speech feature and emotion feature was designed. The recognized specific features of speech and emotion features are extracted by the system with the speech and emotion features extraction module,and classified in model library. The extracted feature information is translated in translation module. The translation module is composed of user monitor,kernel monitor,CPU simulator,instruction translator,peripheral simulator and translation processor,which can realize the accurate translation of translation target. The speech feature recognition module can realize the accurate identification and detection of the system speech and emotion features. The accurate translation results are obtained with translation processor. The experimental results show that the system has high translation accuracy,perfect translation effect,and short system response time.

关 键 词：语音特征 情感特征 翻译系统 功能实现 特征识别 翻译模块 翻译处理器 准确率 

分 类 号：TP391.2[自动化与计算机技术—计算机应用技术]