DISCRIMINATION AND IDENTIFICATION OF AUDITORY PATTERNS

该项目继续进行一系列关于歧视和 识别复杂的听觉模式。这样做的一般目的是 工作是确定人类听众提取 来自复杂声音的信息，特别是那些具有时态的声音 和接近语音的频谱特性。实验使用 受试者受训的标准控制的心理物理方法 直到接近渐近性能在各种区分和 身份识别任务。单个音素长度的声调成分 字长模式基本上可以像在 这些组件是单独呈现的，但仅当 最小不确定度测试程序允许听众将注意力集中在 特定组件。在更高水平的刺激不确定性下，确定 模式的受限光谱-时间区域(晚期、高频 组件)仍然被准确地解析，而大多数其他组件不被准确地解析。 测量包括：选择性的光谱和时间范围 听觉注意，听觉知觉学习的时间进程， 模式的总体属性可以作为线索的方式 有选择地关注特定的时间部分，信息限制 模式辨别和听者学习注意的能力 单个图案内的多个组件，并识别小的 模式的“词汇表”。这项工作的范围现在正在扩大。 包括对听觉模式中个体差异的研究 正常和受损听者之间的辨别能力，以及在 有某些与语言有关的残疾的人群。听觉模式 由合成的声音序列组成，以结合源和 此外，还将研究人类声道的共振特性。 整个系列实验的理论意义在于 它提供了基于人类听者的听觉理论之间的联系 检测和辨别孤立音调的能力和理论 言语感知。这项工作对残疾人具有现实意义。 听众，对他们来说，仅有敏感性措施是不够的 助听器有望带来帮助的预测者。最后，使用 新设计的具有学习和识别功能的听觉辨别测试组件 语言障碍的听众可能有助于解决可能的 这些障碍中的听觉成分。

This project continues a series of experiments on the discrimination and identification of complex auditory patterns. The general purpose of this work is to determine the limits of human listeners' abilities to extract information from complex sounds, including especially those with temporal and spectral properties approximating speech. Experiments use criterion-controlled psychophysical methods in which listeners are trained until approaching asymptotic performance in various discrimination and identification tasks. Individual phoneme-length tonal components of word-length patterns can be resolved essentially as accurately as when those components are presented in isolation, but only when minimal-uncertainty testing procedures permit listeners to focus on a specific component. Under higher levels of stimulus uncertainty certain restricted spectral-temporal regions of the patterns (late, high-frequency components) are still resolved accurately, while most others are not. Measurements include: The spectral and temporal range of selective auditory attention, the time course of auditory perceptual learning, the manner in which gross properties of patterns can serve as cues to selectively attend to specific temporal components, informational limits on pattern discrimination, and listeners' abilities to learn to attend to multiple components within single patterns and to identify small "vocabularies" of patterns. The scope of this work is now being increased to include studies of individual differences in auditory pattern discrimination abilities among normal and impaired listeners, and in populations with certain language-related disabilities. Auditory patterns consisting of sequences of sounds synthesized to incorporate the source and resonant properties of the human vocal tract will also be investigated. The theoretical significance of the overall series of experiments is that it provides a link between auditory theories based on human listeners' abilities to detect and discriminate among isolated tones and theories of speech perception. This work has practical significance for impaired listeners, for whom sensitivity measures alone have been inadequate predictors of the help to be expected from hearing aids. Finally, use of a newly designed auditory discrimination test battery with learning and language-disordered listeners may help to resolve the issue of the possible auditory component in those disorders.