基于多传感器融合的语音数据库构建与应用  

Construction and application of speech database based on multi-sensor fusion

作　　者：陈倩倩[1] CHEN Qianqian(Xi’an Siyuan University,Xi’an 710038,China)

机构地区：[1]西安思源学院,西安710038

出　　处：《自动化与仪器仪表》2025年第3期144-148,共5页Automation & Instrumentation

基　　金：陕西省“十四五”教育科学规划2023年度课题《基于市场需求的国际中文教育专业人才培养体系建构研究》(SGH23Y2875);西安思源学院2023年度校级科研项目《基于市场化发展需求的国际中文教育专业人才培养体系建构研究》(XASYZX-2207);教育部中外语言交流合作中心2023年国际中文教育研究课题《高校国际中文教师职业发展路径研究》(23YH13D)。

摘　　要：为了提高中文语音数据库中语音识别的准确率、扩大语音数据存储量,研究提出使用多传感器融合技术构建关于中文的语音数据库。为验证多传感器融合技术的性能,研究对该技术进行了模拟实验,结果显示,多传感器融合技术数据融合的准确率为95.8%,对语音数据的查全率和查准率分别为98.9%和99.8%。而对基于多传感器融合的中文语音数据库进行实证分析,结果显示,该语音数据库中的语音音高、音调、音长、音质的准确率分别为97%、96%、94%和97%。由上述结果可知,研究提出的基于多传感器融合的中文语音数据库能够提高语音识别的准确率,扩大数据库容量。In order to improve the accuracy of speech recognition in Chinese speech databases and expand the storage capacity of speech data,it is proposed to use multi-sensor fusion technology to construct a speech database about Chinese.To verify the performance of multi-sensor fusion technology,simulation experiments were conducted on the technology.The results showed that the accuracy of data fusion using multi-sensor fusion technology was 95.8%,and the recall and precision of speech data were 98.9% and 99.8%,respectively.Empirical analysis was conducted on the Chinese speech database based on multi-sensor fusion,and the results showed that the accuracy rates of pitch,pitch,length,and sound quality in the speech database were 97%,96%,94%,and 97%,respectively.From the above results,it can be concluded that the proposed Chinese speech database based on multi-sensor fusion can improve the accuracy of speech recognition and expand the database capacity.

关 键 词：语音数据库 数据融合 信息融合 多传感器融合 语音识别 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]