Optimization of a slow-acting automatic gain control system for use in hearing aids

自动增益控制 瞬态(计算机编程) 计算机科学 噪音(视频) 组分(热力学) 控制理论(社会学) 语音识别 电信 控制(管理) 物理 带宽(计算) 人工智能 放大器 热力学 操作系统 图像(数学)
作者
Brian C. J. Moore,Brian R. Glasberg,Michael A. Stone
出处
期刊:British Journal of Audiology [Informa]
卷期号:25 (3): 171-182 被引量:61
标识
DOI:10.3109/03005369109079851
摘要

This paper describes experiments evaluating and optimizing an automatic gain control system, called dual front-end AGC (abbreviated as D), intended for use in hearing aids. This system has two purposes: (1) to compensate for variations in the overall level of speech from one situation to another by slowly changing its gain; (2) to protect the user from sudden intense transients without affecting the long-term gain. This is achieved by using two control voltages to determine the gain. One changes slowly as the input varies in level. Normally this component determines the overall gain. The other comes into operation when an intense transient occurs. It acts rapidly to reduce the gain, avoiding over-amplification of the transient, but its action ceases quickly after the end of the transient. We describe four experiments measuring speech intelligibility for subjects with cochlear hearing loss in which we determine optimum values for two of the time constants of the D system, namely the recovery time of the fast component and the attack time of the slow component. The experiments also compare the D system with linear amplification (L) and 'adaptive compression' (A). The results show: (1) for the D system, optimum values are about 80-150 ms for the recovery time of the fast component and 150-325 ms for the attack time of the slow component; (2) in situations where intense transient sounds are present, and there is either no background sound (experiment 1) or continuous speech-shaped noise as a background (experiment 2), the D system gives significantly better performance than the L or A systems. When the background noise is a single voice, reversed in time (experiment 3), the D and L systems give similar performance, and both are markedly superior to the A system; (3) when the level of speech is varied over a range of 30 dB (experiment 4), both D and A systems allow good performance over the whole range of levels. Performance for the L system worsens markedly at the lower levels.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
王睿哲发布了新的文献求助10
1秒前
嗯嗯发布了新的文献求助10
1秒前
Ava应助懂冬冬采纳,获得10
2秒前
bai发布了新的文献求助10
3秒前
鑫鑫应助沉默的钵钵鸡采纳,获得10
3秒前
领导范儿应助小小牛马采纳,获得10
3秒前
Danny完成签到,获得积分20
4秒前
doranlou发布了新的文献求助30
4秒前
早川木槿完成签到,获得积分10
4秒前
Huangyoko完成签到,获得积分10
4秒前
AllenFeng完成签到,获得积分20
5秒前
6秒前
Song完成签到 ,获得积分10
6秒前
英俊的铭应助Jilo采纳,获得10
6秒前
Nong完成签到,获得积分10
6秒前
Nik- KC完成签到,获得积分10
7秒前
充电宝应助岳约翰采纳,获得10
8秒前
慕青应助科研通管家采纳,获得10
8秒前
aaaa应助科研通管家采纳,获得10
8秒前
NexusExplorer应助chen采纳,获得10
9秒前
9秒前
完美世界应助阳光灿烂采纳,获得10
9秒前
猪猪侠完成签到,获得积分10
10秒前
10秒前
在水一方应助科研通管家采纳,获得10
11秒前
科研通AI2S应助科研通管家采纳,获得10
11秒前
情怀应助科研通管家采纳,获得10
11秒前
牧青应助科研通管家采纳,获得30
11秒前
大个应助科研通管家采纳,获得30
11秒前
生动盼兰完成签到,获得积分10
11秒前
11秒前
12秒前
Lucas应助科研通管家采纳,获得10
12秒前
今后应助科研通管家采纳,获得10
12秒前
zsp发布了新的文献求助10
12秒前
12秒前
12秒前
迅速采梦发布了新的文献求助10
12秒前
12秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
China Pluperfect I: Epistemology of Past and Outside in Chinese Art 520
Management and the Arts 510
Matrix Methods in Data Mining and Pattern Recognition Second Edition 510
基于锂离子电池正极材料回收的绿色溶剂开发及工程化应用研究 500
Auslegungsgeschichte 500
Transdermal drug delivery systems market size report 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 工程类 有机化学 化学工程 生物化学 计算机科学 内科学 物理 复合材料 催化作用 细胞生物学 无机化学 光电子学 物理化学 电极 基因
热门帖子
关注 科研通微信公众号,转发送积分 7641545
求助须知:如何正确求助?哪些是违规求助? 9214625
关于积分的说明 19766578
捐赠科研通 7206987
什么是DOI,文献DOI怎么找? 3276254
关于科研通互助平台的介绍 2437981
邀请新用户注册赠送积分活动 2273852