动物听觉研究

Study on Animal Auditory System

本研究以动物行为和动物神经生理为对象，重点探讨如下四个问题：听觉优先效应的神经生物学机制；信息掩蔽和能量掩蔽的神经生物学机制；听觉认知过程与感觉门控过程之间的交互作用；听觉情绪学习的神经生物学机制。

In studying animal behavior and neurophysiology, this topic concentrates on hearing mechanisms, including the neural mechanism s of auditory precedence effect, the neural mechanism s of informational masking and energetic masking, the interaction between auditory cognitive processes and sensory gating processes, and the neurobiological mechanisms of auditory emotional learning.

动物行为

Animal behavior

动物行为实验主要以听觉惊反射的前脉冲抑制作为听觉感觉门控模型，并以条件化恐惧作用作为情绪学习 模型，研究听知觉组织和感觉门控、情绪、学习、记忆的交互作用。

The study of the interaction between the auditory perceptual organization and sensory gating, emotion, learning, and memory is based on animal experiments, which use the prepulse inhibition of auditory startle reflex as the auditory gating model, and use the fearing conditioning as the emotional learning model .

大鼠惊反射装置

Rat's startle reflex equipment

动物神经生理

Animal neurophysiology

动物神经生理实验主要以 TDT- Ⅱ 系统 (Tucker-Davis Technologies, FL, USA) 在屏蔽箱或隔音室中记录 听觉神经系统场电位，研究优先效应，空间分离释放信息掩蔽和能量掩蔽等神经生理现象。

Animal neurophysiological experiments record neural field potentials in the shielded chamber or sound attenuating chamber by TDT- Ⅱ system, trying to find the neural correlates of the precedence effect and the releasing effect of perceived spatial separation on informational masking and energetic masking.

电磁屏蔽箱

Electromagnetic shielded chamber

动物神经生理实验室

Animal neurophysiological lab

头相关传输函数的建模与测量

Modeling & Measurement of Head-Related Transfer Functions

自由场中声源所发声波到达双耳这一通路的传输函数称为头相关传输函数。它在双耳听觉定位和虚拟三维声音等研究和应用中，具有重要意义。

Head-Related Transfer Functions (HRTFs) are defined as the frequency responses from the sound source to both ears after filtered by the head, torso and pinna in the free field. How to obtain HRTFs has always been a hotspot in psychoacoustics, electroacoustics and related researches. HRTFs play important roles in binaural hearing researches and virtual sound applications.

头相关传输函数与声源的距离、水平方位角、仰角、频率以及个 体形状有关。本研究从解析方法、数值计算和实际测量三个方面展开。

HRTFs vary with the location (distance, azimuth, elevation) of the sound source, frequency and subject. The group involves in the analytical modeling, numerical computation and measurements.

解析方法Analytical Modeling

把头和躯干看作单球体或多球体的组合，利用散射理论和多极重展开的方法在自由声场中进行求解，得到需求点声压，从而获得相应的传输函数。

The head and torso can be modeled as a single sphere or a combination of spheres, the scattering theory and multipole re-expansion method are applied to solve the scattering problem in the free field, by which HRTFs can be obtained.

数值计算Numerical Computation

对头和躯干进行简化或直接扫描，得到三维模型，进行网格划分，再利用数值计算（如边界元、时域有限差分等）求解 声压，从而获得相应的传输函数。

A 3-D model was generated by predigesting or directly scanning the head and torso. Then the model is meshed and solved the problem using numerical computation such as BEM (Boundary Element Method) and FDTD (Finite-Difference Time-Domain method).

实际测量Measurement

以往在实际测量头相关传输函数时，缺乏近场数据。为了满足近场情况下“点声源”的要求，本研究提出了电火花冲激点声源的测量方法，成功地获得了近场数据。

Previous HRTF measurement were lack of near field data. Considering the point source requirement, the electric spark which can generate impulse -like signals is employed in this measurement and the near field HRTF data are obtained efficiently.

测量环境

Measurement Environment

汉语英语听感知差异及言语可懂度

The differences of auditory perception between Chinese and English & Speechintelligibility

汉语和英语在语音学和语言学上的差异决定了它们在听感知上的差异。本研究重点探讨知觉性空间分离对信息掩蔽的释放作用，语音信息相似性与信息掩蔽之间的关系。进而开展言语可懂度理论研究，包括能量掩蔽和信息掩蔽计算模型 ， 以及基于听感知组织的言语可懂度计算模型。

The linguistic and phonetic disparities between Chinese and English lead to differences of auditory perception. We focused on the difference of auditory perception between Chinese and English, and studied the releasing effect of perceived spatial separation on informational masking, the relationship between the speech similarity and informational masking. We also tried to design computational models of speech intelligibility based on the previous psychoacoustic experiments.

汉语的言语识别中，知觉性空间分离对噪声掩蔽的释放约为1dB，对言语掩蔽的释放约为3dB。与英语的对应研究对比发现：知觉性空间分离对汉语中信息掩蔽的释放小于英语。

The experiment al results revealed that the releasing effect of perceived spatial separation in Chinese speech recognition is about 1 dB on noise masking and is about 3 dB on speech masking , which is less than that in English.

本研究得到了 国家自然科学基金重点项目《汉语和英语听感知差异及其应用研究》，教育部“新世纪优秀 人才”计划项目《汉语英语可懂度差异研究》 的支持。

The research is supported by the Key Project of Natural Science Foundation of China and the New Century Excellent Researcher Award Program from Ministry of Education of China.

知觉性空间分离对信息掩蔽的释放作用

The releasing effect of perceived spatial separation on informational masking of Chinese

心理声学实验室

Psychoacoustics Lab

TDT 系统 III

TDT system III