Neural Investigations of Face Perception and Attention Using Population Receptive Field Modeling

使用群体感受野模型进行面部感知和注意力的神经研究

• 批准号: 1756035
• 负责人: Kalanit Grill-Spector
• 金额: $ 61.24万
• 依托单位: Stanford University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-02-15 至 2022-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1756035&HistoricalAwards=false
• 关键词: Neural; Investigations; Face; Perception; Attention

Every day, our many social interactions with others depend on our ability to recognize people from their faces. However, about 2% of the general population have deficits in face perception. What makes some people good at recognizing faces and others bad at it? Scientists have identified brain regions involved in perceiving faces. Additionally, behavioral research has shown that face perception is "holistic". That is, perception of the whole face, all at once, is better than that of individual facial features. However, it is unknown what neural computations underlie this holistic perception of faces. Because holistic face processing requires spatial integration across face parts, the goal of the research is to examine whether such processing relies on spatial integration by neurons in face-selective brain. Additionally, the research will determine if individual differences in neuronal spatial integration predict individual differences in face recognition ability. This project is expected to impact society in several ways. Understanding the neural mechanisms of face recognition could improve measurement, enhance diagnosis, and ultimately improve the wellbeing of individuals with deficits in face recognition. The project will also develop research infrastructure by making publicly available the new methods and computational tools for measurement and modeling of brain responses. The project intends to advance the participation and education of women in STEM fields.The research program addresses important gaps in knowledge by leveraging cutting-edge neuroimaging and computational approaches, including population receptive field (pRF) modeling and hierarchical convolution neural networks, together with behavioral measurements of face perception. In particular, the pRF approach offers a unique opportunity to operationalize and measure spatial processing and attentional effects on face perception in a computational framework. Thus, the research (i) will advance understanding of the neural mechanisms of face perception by quantifying spatial aspects of face processing and attention, (ii) will uncover how spatial integration of different face stimuli, under differing attentional demands explains individual differences in face perception abilities, and (iii) will generate an up-to-date and neurally accurate computational model of human face perception.

每天，我们与他人的许多社交互动都取决于我们从面孔上识别人的能力。然而，大约 2% 的普通人群存在面部感知缺陷。是什么导致有些人擅长识别面孔，而另一些人则不擅长？科学家们已经确定了与感知面孔有关的大脑区域。此外，行为研究表明，面部感知是“整体的”。也就是说，对整个面部的一次性感知优于对单个面部特征的感知。然而，目前尚不清楚这种面部整体感知背后的神经计算是什么。由于整体面部处理需要跨面部部分的空间整合，因此该研究的目标是检查这种处理是否依赖于面部选择性大脑中神经元的空间整合。此外，该研究还将确定神经元空间整合的个体差异是否可以预测面部识别能力的个体差异。该项目预计将以多种方式影响社会。了解面部识别的神经机制可以改善测量，增强诊断，并最终改善面部识别缺陷个体的福祉。该项目还将通过公开用于测量和建模大脑反应的新方法和计算工具来开发研究基础设施。该项目旨在促进女性在 STEM 领域的参与和教育。该研究计划利用尖端的神经影像和计算方法，包括群体感受野 (pRF) 建模和分层卷积神经网络，以及面部感知的行为测量，来解决重要的知识差距。特别是，pRF 方法提供了一个独特的机会，可以在计算框架中操作和测量空间处理以及注意力对面部感知的影响。因此，该研究（i）将通过量化面部处理和注意力的空间方面来促进对面部感知神经机制的理解，（ii）将揭示不同注意力需求下不同面部刺激的空间整合如何解释面部感知能力的个体差异，以及（iii）将生成最新的、神经精确的人脸感知计算模型。

项目成果

Holistic face recognition is an emergent phenomenon of spatial processing in face-selective regions.

• DOI: 10.1038/s41467-021-24806-1
• 发表时间: 2021-08-06
• 期刊: Nature communications
• 影响因子: 16.6
• 作者: Poltoratski S;Kay K;Finzi D;Grill-Spector K
• 通讯作者: Grill-Spector K

Using population receptive field models to elucidate spatial integration in high-level visual cortex

使用群体感受野模型来阐明高级视觉皮层的空间整合

• DOI: 10.32470/ccn.2019.1178-0
• 发表时间: 2019
• 期刊: Germany
• 作者: Poltoratski, Sonia Kay