基于传感器技术的三维发音可视化合成系统研究设计  

Research and design of 3D articulation visualisation and synthesis system based on sensor technology

作　　者：吴亚亚[1] WU Yaya(Shangluo Vocational and Technical College,Shangluo Shaanxi 726000,China)

机构地区：[1]商洛职业技术学院,陕西商洛726000

出　　处：《自动化与仪器仪表》2024年第4期102-105,共4页Automation & Instrumentation

基　　金：2023年商洛职业技术学院一般课题《基于省级精品在线开放课程建设的PBL项目式学习研究——以<现代汉语>为例》(JYKT202318)。

摘　　要：对于汉语非母语学习者,正确的发音是一个难题,而传统的发音教学方法难以直观地展示发音的运动过程。因此,这项研究通过将传感器技术应用于发音过程的捕捉和转化,使学习者能够直观地观察嘴唇和舌头的运动,从而更好地模仿和改进发音。实验结果表明,传感器采集的音频信号噪声干扰小且质量良好,输出的电信号与声压之间存在良好的线性关系。实验系统在声音信号和喉颈运动的识别上方面表现良好,准确度超过80%,最后成功地实现了可视化合成效果,应用该方法的教学班级的发音准确率最高达到了90%。总之,基于传感器技术的三维发音可视化合成系统为发音研究和教学提供了一种有效的工具和方法,有望在语音学和教育领域产生积极的影响。For non-native learners of Chinese,correct pronunciation is a difficult task,and traditional pronunciation teaching methods are difficult to visualize the movement process of pronunciation.Therefore,this study enables learners to visualize the movements of lips and tongue by applying sensor technology to capture and transform the pronunciation process,so that they can better imitate and improve pronunciation.The experimental results show that the audio signals acquired by the sensor have low noise interference and good quality,and there is a good linear relationship between the output electrical signal and the sound pressure.The experimental system performs well in terms of recognition of sound signals and laryngeal and neck movements with an accuracy of more than 80%,and finally the visual synthesis effect is successfully achieved,and the pronunciation accuracy of the teaching class applying the method reaches up to 0.9.In conclusion,the 3D pronunciation visual synthesis system based on the sensor technology provides an effective tool and methodology for pronunciation research and teaching,and is expected to have a have a positive impact.

关 键 词：声音传感器 发音识别 三维可视化 合成系统 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]