Contrast-based face adaptation: a new method for exploring the organization of face-space

基于对比度的面部适应：探索面部空间组织的新方法

• 批准号: 355879-2008
• 负责人: Barton, Jason
• 金额: $ 2.04万
• 依托单位: University of British Columbia
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2010
• 资助国家: 加拿大
• 起止时间: 2010-01-01 至 2011-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=459616
• 关键词: Contrast; based; face; adaptation; new

Faces are some of the most complex visual stimuli that the human brain processes. An important question in research on object recognition is how the visual system organizes and accesses neural representations of faces. Adaptation is a powerful experimental technique: by examining how responses to different faces are affected by preview of a particular face, we can understand what the system encodes about faces, how these are represented, and how these representations interact. We have developed a new adaptation technique that asks how the amount of contrast needed to perceive a specific face is affected by preview of the same or other faces. Our technique has shown that a short preview facilitates perception of the same face, but inhibits perception of other faces. We can use this effect to ask a series of key questions. Does the degree of inhibition depend upon the similarity of the other face to the adapting face? Is there a distinct segregation of face space by gender? How do representations of the identity of the face and representations of facial expression interact? Is the organization of these representations dependent on face-expert mechanisms that are not available when faces are seen upside-down? Our plan is to test the hypotheses driving these questions and use computational models of neural networks to simulate the findings and to suggest how the neural representations of these complex visual stimuli might be organized and interact. This type of knowledge will advance our understanding of the storage of object information and the mechanisms that contribute to the rapid and effortless recognition performance of the human visual system.

面孔是人类大脑处理的最复杂的视觉刺激之一。物体识别研究中的一个重要问题是视觉系统如何组织和访问人脸的神经表征。适应是一种强大的实验技术：通过研究对不同面孔的反应如何受到特定面孔的预览的影响，我们可以了解系统对面孔编码的内容，这些内容是如何表示的，以及这些表示如何相互作用。我们开发了一种新的适应技术，该技术询问感知特定面部所需的对比度如何受到相同或其他面部预览的影响。我们的技术表明，短暂的预览有助于感知同一张脸，但抑制感知其他面孔。我们可以利用这种效应提出一系列关键问题。抑制的程度是否取决于另一张脸与适应脸的相似性？面部空间是否有明显的性别隔离？面部身份表征和面部表情表征是如何相互作用的？这些表征的组织是否依赖于面部专家机制，而当面部被颠倒时，这种机制是不可用的？我们的计划是测试驱动这些问题的假设，并使用神经网络的计算模型来模拟研究结果，并提出这些复杂视觉刺激的神经表征可能是如何组织和相互作用的。这种类型的知识将促进我们对物体信息的存储以及有助于人类视觉系统快速轻松识别性能的机制的理解。