Speech Pattern Recognition Processes with Normal and Cochlear-Implant Produced Hearing

正常听力和人工耳蜗产生的听力的语音模式识别过程

• 批准号: 8808602
• 负责人: Vivien Tartter
• 金额: $ 0.06万
• 依托单位: Rutgers University New Brunswick
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1989
• 资助国家: 美国
• 起止时间: 1989-01-01 至 1989-05-01
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=8808602&HistoricalAwards=false
• 关键词: Speech; Pattern; Recognition; Processes; Normal

A significant recent advance in medical technology has been the development of the cochlear implant. This is a device which is surgically attached to the cochlea, or inner ear, connecting electrodes to cochlear nerves. The wearer carries a small radio- like device outside the body which translates sounds into elec- trical impulses transmitted directly to cochlear nerves via the electrodes, and the impulses are "heard" by the brain in a way which the wearer is able to interpret. For people with certain kinds of deafness, cochlear implants have already provided significant enhancements in their ability to understand speech, most usually in conjunction with lip reading but sometimes even in the absence of visual cues. The growing success of cochlear implantation presents interesting challenges to the science of speech perception, since only a fraction of the information in the acoustical signal is actually translated into electrical impulses. To the extent that the user is able to interpret the impoverished signal accurately, this shows that key perceptual information is included in it, thus affording a new perspective on the fundamental problem of speech perception: what are the acoustic properties of speech which determine accurate perception? This research project is a systematic exploration of the nature of the signal received by cochlear implantees, and their perception of it. The research will focus on wearers of the newly developed "Australian nucleus" implant, which is a multi- channel device and carries much more information than earlier models. The researcher will examine the speech perception of implantees, the changes in their speech patterns over time, and the reactions of normal listeners to speech signals distorted in the ways which the implant device distorts them for its users. The principal research aim of the project is to identify the components of the distorted signal which enable correct percep- tion. A successful outcome will be an important contribution to speech science, as well as offering the possibility of further improvements in the design of cochlear implants.

最近医学技术的一个重大进步是人工耳蜗术的发展。这是一种通过手术将电极连接到耳蜗神经或内耳上的装置。佩戴者在体外携带一个类似无线电的小型设备，将声音转化为电子脉冲，通过电极直接传输到耳蜗神经，大脑以佩戴者能够解释的方式“听到”这些脉冲。对于某些类型的耳聋患者，人工耳蜗已经显著增强了他们理解语言的能力，大多数情况下是与唇读相结合，但有时甚至在没有视觉提示的情况下也是如此。人工耳蜗植入的日益成功对语音感知科学提出了有趣的挑战，因为实际上只有一小部分声音信号中的信息被转化为电脉冲。在用户能够准确地解释贫乏的信号的程度上，这表明其中包含了关键的感知信息，从而为语音感知的基本问题提供了一个新的视角：决定准确感知的语音的声学特性是什么？本研究项目是对人工耳蜗者接收到的信号的性质以及他们对信号的感知的系统探索。这项研究将集中在新开发的“澳大利亚核”植入物的佩戴者身上，这是一种多通道设备，比以前的型号携带的信息要多得多。研究人员将检查植入者的言语感知，他们的言语模式随时间的变化，以及正常听者对植入设备为用户扭曲的语音信号的反应。该项目的主要研究目标是识别失真信号的成分，使其能够正确感知。一个成功的结果将是对语音科学的重要贡献，并为进一步改进人工耳蜗的设计提供了可能性。