Predicting listeners' attentional bias toward the human voice: perceptual, neural, and semantic factors.

想象一下，一个女孩在操场上玩耍：她听到风在树上呼啸，汽车呼啸而过，另一个孩子和她说话的声音，还有电子玩具发出的音乐。她可以注意到这些声音中的任何一个，然而她的注意力很可能很快被人类的声音所吸引。人类每天最常见的任务之一就是有选择地注意与交流相关的声音，而积极地忽略在我们忙碌的听觉环境中如此普遍的其他声音。这是一项非常复杂的工作，尤其是对正在成长的听众来说。然而，我之前的研究发现，婴儿、儿童和成人在复杂的场景中对言语的注意力存在持续的偏见。我的研究计划旨在揭示，当许多其他声音争夺注意力时，所有年龄段的听众都倾向于在忙碌的声学环境中听演讲。这项工作的目标是描述从婴儿到成年的真实世界场景中人类交流的感知，认知和神经构建块。 尽管所有年龄段的听众在复杂场景中都表现出注意力言语偏差，但由于认知能力的迅速发展，从婴儿期到成年期，将注意力转向言语的因素可能有所不同。使用一个两阶段的方法，我首先提出的行为研究，在幼儿和成年期的注意力言语偏见的声学和语义驱动程序的特点。其次，利用脑电无创神经成像技术，探索了复杂场景下言语和非言语声音注意的神经指标。这个NSERC DG提案将通过以下问题揭示所有年龄段的听众是如何偏向于在复杂场景中听演讲的：1。什么是注意言语偏差？是偏向于某些类型的显著声学特征，还是偏向于语义上有意义的信号？ 2.注意言语偏差的驱动因素在儿童期到成年期是否会因为婴儿和学龄前儿童的知觉和认知发展而发生变化？ 3.在忙碌的场景中，特别是在婴儿期，神经测量能否用于索引对语音和非语音声音的注意力？ 深入了解我们在忙碌的听力环境中倾听演讲的非凡能力，将有助于为所有年龄段的听众制定早期有效的干预措施，这些听众在忙碌的场景中难以理解演讲。我们将开发创新的神经方法，为婴儿的听觉体验提供一个窗口，并将表征复杂场景中真实的世界声音的典型注意力层次。这些发现可以在检测变化可能是使命关键的工作场所（例如武装部队）中利用。这份提案中概述的行为和神经研究将使基础科学更接近于理解所有年龄段的人类如何在真实的世界中成功地交流。

Picture a girl playing on the playground: She hears the roar of wind in the trees, the whirr of cars driving by, the voice of another child talking to her, and music emanating from an electronic toy. She could attend to any one of these sounds, and yet her attention is likely to be quickly grasped by the human voice. One of the most common tasks humans do on a daily basis is selectively attend to sounds that are relevant for communication, while actively ignoring other sounds that are so prevalent in our busy auditory environments. This is a surprisingly complex job, especially for developing listeners, and yet my previous research uncovered a persistent attentional bias toward speech in complex scenes for infants, children, and adults. My research program is designed to uncover how it is that listeners of all ages are biased to attend to speech in busy acoustic environments, when many other sounds compete for attention. The goal of this work is to characterize the perceptual, cognitive, and neural building blocks of human communication in real-world scenarios from infancy to adulthood. Although listeners of all ages show an attentional speech bias in complex scenes, the factors that drive attention toward speech from infancy to adulthood may differ due to burgeoning cognitive abilities. Using a two-stage approach, I first propose behavioural studies that characterize acoustic and semantic drivers of the attentional speech bias in early childhood and adulthood. Second, I explore a neural index of attention to speech and non-speech sounds in complex scenes using non-the invasive neuroimaging technique of EEG. This NSERC DG proposal will uncover how listeners of all ages are biased to attend to speech in complex scenes through the following questions: 1. What is the attentional speech bias? A bias toward certain types of salient acoustic characteristics or a bias toward semantically meaningful signals? 2. Do the drivers of the attentional speech bias change in childhood to adulthood because of perceptual and cognitive development in infants and preschoolers? 3. Can neural measures be used to index attention to speech and non-speech sounds in busy scenes, especially in infancy? An in-depth understanding of our remarkable ability to attend to speech in busy listening settings will lead to the development of early and effective interventions for listeners of all ages that struggle to comprehend speech in busy scenes. We will develop innovative neural methods that provide a window into the infant listening experience and will characterize the typical attentional hierarchy for real world sounds in complex scenes. These findings can be leveraged in workplaces where detecting a change can be mission critical, such as the Armed Forces. Together the behavioural and neural studies outlined in this proposal will move basic science one step closer to understanding how humans of all ages successfully communicate in the real world.