论文部分内容阅读
目前电子耳蜗在汉语的应用环境下会丢失音调信息,导致言语识别率的降低。为了便于针对性地研究适合汉语特征的电子耳蜗言语处理算法并提高言语识别率,本文在改进的电子耳蜗前端信号采集平台上研究了不同电子耳蜗言语处理策略的波形特性以及语谱分布、能量强度、基频和共振峰等谱相关参数,通过分析和提取参数特征,研究了两大类电子耳蜗言语处理策略的特征。因此,本文的研究目的是期望通过本文结果能够有助于扩展基于汉语特征的电子耳蜗言语处理策略的研究。
At present, the electronic cochlear will lose the pitch information in the Chinese application environment, resulting in a decrease in speech recognition rate. In order to facilitate the targeted study of Chinese cochlear speech processing algorithm and improve speech recognition rate, this paper studied the waveform characteristics and the distribution of speech spectrum of different cochlear speech processing strategies on the improved Cochlear front-end signal acquisition platform. The energy intensity , Fundamental frequency and resonance peak, etc. The characteristics of speech processing strategies in two categories of cochlear implants were studied by analyzing and extracting the parameter characteristics. Therefore, the research purpose of this paper is to expect that the results of this paper can be helpful to expand the study of cochlear speech processing strategies based on Chinese characteristics.