Sound Source Segregation and Determination

声源分离与判定

基本信息

批准号：
7227826
负责人：
RICHARD R FAY
金额：
$ 21.05万
依托单位：
LOYOLA UNIVERSITY OF CHICAGO
依托单位国家：
美国
项目类别：
财政年份：
2003
资助国家：
美国
起止时间：
2003-05-01 至 2010-04-30
项目状态：
已结题

项目摘要

DESCRIPTION (provided by applicant): The most important features of the human sense of hearing are those that promote the segregation and integration of acoustic features in a way that results in the determination of actual sound sources. Since environments often include multiple and simultaneous objects, the auditory system must function fundamentally as a "separation machine" whose job it is to determine individual sources from the complex mixture of source sounds reaching the ears. This sort of auditory scene analysis is now well appreciated as an important function of human hearing. We suggest here that these functions are also primitive features of a general vertebrate sense of hearing, and that these are the functions that have defined success and fitness most importantly throughout the evolution of the vertebrate auditory system. The overall goals are to determine how these primitive functions of the vertebrate sense of hearing operate in a relatively simple vertebrate species. One goal is to help establish a biological and evolutionary context within which human hearing can be more fully understood. Aim 1: We will investigate the capacities of a simple vertebrate species to determine source characteristics when there are multiple, simultaneous sound sources present (i.e., auditory scene analysis and source segregation). We will investigate other acoustic features that promote auditory perceptual segregation in human listeners for their effects on fish listeners. These features include onset asynchrony, amplitude modulation, duration, harmonic structure, (and spatial origin: Aim 2). These features will be investigated for their effects on perception using behavioral conditioning methods using stimulus generalization paradigms. Aim 2: We will investigate auditory source segregation using multiple, discrete local sources. We will analyze the effects of the spatial origin of sources independently of the non-spatial acoustic features (Aim 1) that may play a role in sound source sesegregation. Aim 3: We will determine the nature of the neural codes and representations underlying source segregation by recording from peripheral and central auditory neurons in response to the same signals and sources used in the above behavioral experiments. We will focus on understanding the dimensions of neural activity that the brain could use in segregating and integrating simultaneous acoustic components from one or more sources.

描述（由申请人提供）：人类听觉最重要的特征是那些促进声学特征的分离和整合的特征，从而确定实际声源。由于环境通常包含多个同时存在的对象，因此听觉系统必须从根本上发挥“分离机器”的作用，其工作是从到达耳朵的复杂的源声音混合中确定单个声源。这种听觉场景分析现在被认为是人类听觉的重要功能。我们在这里建议，这些功能也是一般脊椎动物听觉的原始特征，并且这些功能在脊椎动物听觉系统的整个进化过程中最重要地定义了成功和适应性。总体目标是确定脊椎动物听觉的这些原始功能如何在相对简单的脊椎动物物种中发挥作用。一个目标是帮助建立一个可以更充分地理解人类听力的生物和进化背景。目标 1：我们将研究简单脊椎动物物种在存在多个同时存在的声源（即听觉场景分析和声源分离）时确定声源特征的能力。我们将研究其他促进人类听众听觉感知分离的声学特征，以及它们对鱼类听众的影响。这些特征包括起始异步、幅度调制、持续时间、谐波结构（和空间原点：目标 2）。将使用刺激泛化范式的行为调节方法来研究这些特征对感知的影响。目标 2：我们将使用多个离散的本地源来研究听觉源隔离。我们将独立于非空间声学特征（目标 1）分析声源空间起源的影响，这些特征可能在声源隔离中发挥作用。目标 3：我们将通过记录外周和中枢听觉神经元对上述行为实验中使用的相同信号和源的响应，确定源分离背后的神经代码和表征的性质。我们将重点了解大脑可用于分离和整合来自一个或多个来源的同步声学成分的神经活动维度。