摘要
为了获得更加理想的普通话测试效果,针对当前计算机辅助普通话测试系统存在的不足,设计了复杂环境下的计算机辅助普通话测试系统。首先构建计算机辅助普通话测试系统的总体流程;然后重点设计软件子系统,该子系统引入小波变换对采集的普通话测试者语音信号进行预处理,去除外界环境对普通话测试结果的干扰,并引入人工智能技术对预处理后的普通话测试者语音信号进行自动识别,获得普通话测试结果;最后通过仿真实验测试了所提系统的性能,结果表明,相对于其他计算机辅助普通话测试系统,该设计系统的计算机辅助普通话测试精度得到大幅度提升,对外界环境的鲁棒性更强,可以更好地满足普通话测试要求,具有更高的实际应用价值。
In view of the shortcomings of the current computer⁃aided Putonghua test system,a computer⁃aided Putonghua test system in complex environment is designed to obtain better effects of Putonghua test.The overall process of the computer⁃aided Putonghua test system is constructed,and then the software subsystem is designed.The wavelet transform is adopted for the subsystem to preprocess the Putonghua testees′voice signals collected,so as to eliminate the interference of the external environment on the Putonghua test results.The artificial intelligence technology is adopted to automatically recognize the preprocessed voice signals and get the results of test.The performance of the proposed system is tested by simulation experiment.The results show that,in comparison with the other computer⁃aided Putonghua test system,the proposed system has a greatly improved accuracy in the computer⁃aided Putonghua test,and its robustness to the external environment is stronger,which can better meet the requirements of Putonghua test.Therefore,the system has a higher practical application value.
作者
廖俐
LIAO Li(South China Normal University,Guangzhou 510000,China;Guangxi Vocational&Technical Institute of Industry,Nanning 530000,China)
出处
《现代电子技术》
2021年第1期149-152,共4页
Modern Electronics Technique
基金
2017年国家社科基金重点项目(17AYY014)。
关键词
计算机辅助测试系统
总体框架
语音信号预处理
软件子系统
人工智能技术
语音信号特征
computer⁃aid test system
general framework
speech signal preprocessing
software subsystem
artificial intelligence technology
voice signal feature
