Revealing the distribution and dynamics of cortical representations in the visual processing of “stuff”

我们对世界的视觉体验是多方面的。仅通过视觉信息，我们就能同时识别出不同抽象层次的物体（椅子/勺子/猫）（我的猫、宠物、有生命的东西）；我们识别制造物品的材料（木材/金属/毛皮）；推断多感官材料品质（硬/冷/蓬松）；我们自动访问相关的语义概念和启示（“可以坐”，“可以抓”，“可以抚摸”）。视觉神经科学的一个重要挑战是理解产生这些多维表征的计算和神经机制。然而，这一挑战在很大程度上没有得到解决，因为大部分研究都集中在对物体的视觉识别上，而最近的研究主要集中在对彼此独立的材料的视觉识别上，没有考虑到所涉及的计算的合作性质。该项目的目的是填补这一空白，并阐明我们多维视觉表征的计算和神经机制。具体来说，这些实验旨在梳理出哪些图像特征和刺激属性在大脑的何时何地被表征，以及这些表征的结构在不同的大脑区域和时间点之间是如何不同的。这将通过使用先进的计算机渲染方法和心理物理学，结合神经成像（功能性磁共振成像和脑磁图）和多变量分析技术，对物体和材料的感知和神经处理进行交叉调查来实现。

Our visual experience of the world is multifaceted. Through visual information alone we simultaneously identify objects (chair /spoon / cat) at different levels of abstraction (my cat, pet, animate living thing); we recognise the materials that things are made from (wood / metal /fur); infer multisensory material qualities (hard / cold / fluffy); and we automatically access associated semantic concepts and affordances (“can sit on”, “can grasp”, “can stroke”). An important challenge in visual neuroscience is understanding the computations and neural mechanisms that give rise to these multidimensional representations. Yet, this challenge has mostly gone unmet, since much of the research investigating representations in the brain has focused on the visual recognition of objects and – more recently – materials largely independently from one another, failing to take into account the cooperative nature of the computations involved. The aim of the proposed project is to fill this gap and to elucidate the computations and neural mechanisms underlying our multidimensional visual representations. Specifically, the experiments aim to tease apart what image features and stimulus properties are represented where and when in the brain, and how the structure of these representations differs between different brain regions and time points. This will be achieved by intersecting investigations into the perception and neural processing of objects and materials using advanced computer-rendering methods and psychophysics, in combination with neuroimaging (functional Magnetic Resonance Imaging and Magnetoencephalography), and multivariate analysis techniques.