A biologically inspired model adding binaural aspects to soundscape analysis
A biologically inspired model adding binaural aspects to soundscape analysis
复制标题
受生物学启发的模型，将双耳方面添加到音景分析中
DOI：
--
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
D. Botteldooren
中科院分区：
文献类型：
--
作者：
Michiel Boes;B. D. Coensel;Damiano Oldoni;D. Botteldooren
双耳听觉是人类听觉系统的基本特性。它为大脑提供有关声音位置的重要信息，这些信息有助于听觉流分离的过程。因此，用于音景的计算分析的模型将受益于双耳分量的添加。本文提出了一种生物学启发的双耳分析模型。模拟了人脑中计算双耳时间差（ITD）和双耳强度差（IID）这两个基本双耳线索的神经通路。因此，神经映射被用来将这些ITD敏感和IID敏感神经元与表示特定方向的神经元链接，以便获得声源方向的估计。这张图在生物学上是合理的，也是联合收割机结合IID和ITD信息的一种优雅的方式。该模型将学习和适应定位以及其他几个功能的人类双耳听力。此外，提出了一个分析模型的有效性，在检测声源的方向，也在分散注意力的声音或背景噪声的存在。
Binaural hearing is an essential property of the human auditory system. It is providing the brain with vital information about the location of sounds and this information contributes to the process of auditory stream segregation. Thus, models for the computational analysis of soundscapes would benefit from the addition of a binaural component. In this paper, a biologically inspired binaural analysis model is presented. The neural pathways in the human brain calculating interaural time difference (ITD) and interaural intensity difference (IID), two essential binaural cues, are modelled. Consequently, a neural map is used to link these ITD-sensitive and IID-sensitive neurons to neurons representing a specific direction, in order to obtain an estimate of sound source direction. This map turns out to be both a biologically plausible and an elegant way to combine IID and ITD information. The model is shown to incorporate learning and adaptation in localization as well as several other features of human binaural hearing. In addition, an analysis is presented of the model's effectiveness in detecting sound source directions, also in the presence of distracting sounds or background noise.