研究方向
  位置: 首页 > 科学研究 > 研究方向 > 正文

听力损伤客观评价方法及听力康复的语音处理策略

作者:   时间:2019-06-19   来源:   点击数:

听力损伤客观评价方法及听力康复的语音处理策略

Objective hearing loss evaluation and speech processing strategies

for hearing rehabilitation

听觉返回抑制 :基于位置和频率维度的ERP研究

Auditory Inhibition of Return – Location and Frequency based ERP

该课题采用听觉事件相关电位的方法研究在听觉通道内基于位置和基于频率的返回抑制现象,结果表明: 听觉通道基于位置和基于频率的返回抑制都是注意调节的结果;基于位置的返回抑制是多个感知觉加工阶段受注意调节而产生的,而基于频率的返回抑制主要受到较晚期加工阶段的影响,它们有着不同的机制。

Using the method of the Event Related Potential (ERP) to investigate the location based and frequency based IOR (Inhibition of Return) in the auditory pathway. The results revealed: 1. Both the location based and frequency based IOR are manipulated by attention; 2. The location based IOR is modulated by attention at various stages, however, the frequency based IOR is modulated mainly at the later stage, which suggested that the two auditory IORs are manipulated by different mechanisms.

位置维度不同SOA差异波的脑地形分布图

听觉稳态响应:新的客观听力评估方法

Auditory Steady-state Response: A New Method for Objective Audiology Evaluation

听觉稳态响应是由调频、调幅或者混合调制的声音信号诱发,记录到诱发频率成分与调制速率一致的早期 诱发电位,其多频同步刺激的方法已应用于临床听力评估中。本课题开展了响度感受变化的全频率客观测 量方法和响度重振客观检测方法的研究。

A steady-state evoked response is a repetitive evoked potential whose constituent discrete frequency components remain constant in amplitude and time over an infinitely prolonged time period . And the typical technology of recording SSR- MASTER (Multiple Auditory Steady-state Evoked Responses) has been applied to auditory clinical test. This research with MASTER involves the loudness objective evaluation and recruitment objective testing for hearing-impaired people.

多频正弦调幅刺激诱发的听觉稳态响应


适用于汉语的人工耳蜗编码策略

Cochlear implant coding strategy forChinese

人工耳蜗是治疗重度感音性耳聋的一个重要手段,其语音处理策略直接影响到感知效果。临床中普遍采用 的CIS算法对于母语为汉语的使用者,其言语可懂度明显低于母语为英语的使用者。本研究提出了一种改进 的人工耳蜗处理策略,使得讲汉语的使用者在噪声环境中也能获得较高的言语可懂度。

Cochlear implanting is an important therapy for profound sensori-neural hearing loss . The strategy of speech processing is essential for the rehabilitation effect . The typical strategy CIS (Continuous Interleaved Sampling ) is not satisfactory for Chinese speakers. In this study, an improved approach SPLS (Simulated Phase-Locking Stimulation) is developed and has been proved effective.

算法框架

Framework of SPLS

研究结论Conclusions

SPLS 通过引入相位信息提高了汉语的声调识别率,以及噪声环境下的言语识别率。

The SPLS strategy improved both the Mandarin tone recognition rate and speech recognition rate under noisy environments for Chinese speakers by introducing phase information.

Copyright 版权所有©北京大学智能学院 All Rrights Reserved.

地址:北京市海淀区颐和园路5号

手机版