电子耳蜗CIS言语处理策略参数特征研究  

Research on Parameter Feature of Continuous Interleaved Sampling Strategy for Cochlear Implant

作　　者：陈又圣[1] 王健[1] 薛国伟[1] 张璞[1] 陈培培[1] 

机构地区：[1]深圳信息职业技术学院电子与通信学院,广东深圳518172

出　　处：《深圳信息职业技术学院学报》2017年第3期12-18,共7页Journal of Shenzhen Institute of Information Technology

基　　金：深圳科技计划基础研究(No.JCYJ20160415114250896);广东省自然科学基金博士启动基金(No.2016A030310073);国家自然科学基金(No.81401539)

摘　　要：目前电子耳蜗主流的言语处理策略是基于滤波器组的语音处理算法,而其中大部分算法是基于连续交替采样策略(即CIS算法)的算法结构延伸而成的。连续交替采样策略是一种通过电极阵列连续交替刺激听神经的算法,在特定的时刻只有一个电极在刺激,因而可以避免电极之间的电流扩散而降低算法的性能。本文研究了连续交替采样策略的算法结构,通过编程实现该言语处理算法并进一步提取语音的特征和参数,研究了连续交替采样策略的信号变化特征。Present main speech processing strategies are filter bank based speech processing algorithm, and most algorithms are based on the algorithm structure of continuous interleaved sampling strategy （CIS algorithm）. CIS method uses electrode array to continuously stimulate auditory nerve one by one. At a fixed time, only one electrode are stimulating the auditory nerve, therefore, this algorithm can avoid the current spread between two electrodes and weaken the algorithm performance. In this paper, we research the algorithm structure of continuous interleaved sampling strategy, and programme the speech processing algorithm. Further, we extract the speech feature and parameter, and research the signal changing feature for CIS algorithm.

关 键 词：电子耳蜗 连续交替采样策略 参数 语音特征 

分 类 号：TN912.3[电子电信—通信与信息系统] R764[电子电信—信息与通信工程]