权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

A Study on Mechanism of Processing, Perception and Cognition of Auditory Signal.

听觉信号处理、感知和认知机制的研究。

基本信息

批准号：
63460126
负责人：
USAGAWA Tsuyoshi
金额：
$ 4.86万
依托单位：
Kumamoto University
依托单位国家：
日本
项目类别：
Grant-in-Aid for General Scientific Research (B)
财政年份：
1988
资助国家：
日本
起止时间：
1988 至 1990
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-63460126/
关键词：
Low Pitch Autocorrelation Model Pitch Shift Complex Tone Dither Detection model of Distortion Attention Reproduction of signal assigned in a few bits マスキングランダム性の知覚ディサピッチ弁別信号検出資源配分妨害感

项目摘要

In this project, we measure frequency difference limen (D. L.) when jitter (random fluctuation of frequency components) added to a pure tone and complex tones which have three components with various durations of stimuli. The results show that increment of D.L. induced by jitter for pure tone was rarer than that for complex tones, and in short term region non-jittered stimuli with higher frequency components had smaller D. L. than one with lower frequency components. These results indicate that D. L. could be related to the pitch strength which derived from autocorrelation model. Furthermore the results obtained pitch match experiments using 3 and 6 components complex tones, especially ones with small pitch shift, support autocorrelation model. However, the results with larger pitch shift can not be explained well using this model.The quantization noise (harmonic distortion) and the broad band noise due to dither are evaluated. The results are as follows ; using exponentially decayed signal as test tone, generally the effectiveness of dither are confirmed as the increase of detection level. And using complex tones which have three or five components as test tone, the effects of dither are changed due to the fundamental frequency of test tones and the dither is not always effective in this case. The detection model of distortion based on the masking effect can explain this phenomenon.From the study of attention of human auditory information processing, the hit rate in detection task increases and the rate of false alarm decreases when temporal information of tone occurrence is give.

在本项目中，我们测量频差阈（D。L.）的范围内。当抖动（频率分量的随机波动）被添加到纯音和具有三个分量的具有不同刺激持续时间的复合音时。结果表明，D.L.在短时程区，高频成分的非抖动刺激诱发的D. L.而不是具有较低频率分量的一个。这些结果表明D. L.可能与自相关模型推导出的基音强度有关。此外，对三分量和六分量复声调，特别是小频移复声调的基音匹配实验结果支持自相关模型。然而，该模型不能很好地解释较大的音调偏移的结果。量化噪声（谐波失真）和宽带噪声的抖动进行了评估。结果表明，以指数衰减信号为测试音，抖动的有效性一般随检测电平的增加而得到证实。当测试音为三个或五个分量的复合音时，由于测试音基频的影响，抖动的效果发生了变化，在这种情况下，抖动并不总是有效的。基于掩蔽效应的失真检测模型可以解释这一现象，从人类听觉信息处理的注意研究来看，当给出声调出现的时间信息时，检测任务的命中率增加，虚警率降低。