首页 | 本学科首页   官方微博 | 高级检索  
     检索      

电子耳蜗的选择性频率幅度调制编码算法
引用本文:关添,叶大田.电子耳蜗的选择性频率幅度调制编码算法[J].清华大学学报(自然科学版),2006,46(12):2057-2060.
作者姓名:关添  叶大田
作者单位:清华大学,生物医学工程系,北京,100084;清华大学,生物医学工程系,北京,100084
基金项目:面向21世纪教育振兴行动计划(985计划);粤港科技合作基金
摘    要:为提高电子耳蜗植入者在噪声下的语音识别能力,提出了一种利用频率调制信息的算法。该算法根据人耳听觉机理,增加了频带选择法则。该法则对不同频带提取出的频率调制信息进行了有选择的传递,并在听觉仿真试验中得到验证:该算法既可以比传统幅度信息编码算法更有效地提高电子耳蜗植入者在噪声下的语音识别能力,又降低了不选择频带而全通道传递频率调制信息算法的复杂度(频率调制信息提取的计算量降低了37.5%),减少了内存占用,使之适合在临床上实时实现。

关 键 词:电子耳蜗  频率调制  频带选择法则  语音信号处理
文章编号:1000-0054(2006)12-2057-04
修稿时间:2005年8月24日

Selective frequency amplitude modulation encoding strategy for cochlear implants
GUAN Tian,YE Datian.Selective frequency amplitude modulation encoding strategy for cochlear implants[J].Journal of Tsinghua University(Science and Technology),2006,46(12):2057-2060.
Authors:GUAN Tian  YE Datian
Abstract:A speech processing strategy was developed to enhance speech recognition with cochlear implant users in noisy environments.The strategy enhanced frequency band selection based on human auditory mechanism.Using this principle,the frequency modulation information was encoded only in the selected frequency bands,and tested in acoustic simulation experiments.This strategy remarkably improves recognition in noisy environments compared with the traditional algorithm which only transmits envelope cues.Also the computing in this algorithm is reduced by 37.5% compared with a similar algorithm which transmits frequency modulation information in full channels without frequency band selection.Therefore,the strategy is suitable for real time processing in clinics.
Keywords:cochlear implant(CI)  frequency modulation(FM)  frequency band selection  speech signal processing
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号