报告题目：Visual perception from lab experiments to real world applications
报 告 人：Xu Hong（徐红）教授
Xu Hong（徐红），新加坡南洋理工大学助理教授。本科毕业于北京大学，于芝加哥大学心理学系师从David Bradley获得博士学位，曾在哥伦比亚大学神经科学系师从钱宁从事博士后研究，已在Cerebral Cortex, Journal of Neuroscience , Journal of Neurophysiology等国际知名学术期刊上发表了30余篇高被引学术论文。累计主持了新加坡教育部、交通管理局等共计约1000万人民币的基金项目。主要研究领域：计算视觉、面孔认知、运动认知、视觉适应后效、人因工程。
To fit the mind to the world, our sensory systems have to adjust constantly to the dynamic environment, a process known as adaptation. Visual adaptation occurs at multiple levels along the cortical hierarchy, starting from low-level areas such as the primary visual cortex (V1) on orientation adaptation – the tilt aftereffect – that the perceived orientation is changed after prolonged inspection of another oriented bar or grating, to higher-level areas such as fusiform face area (FFA) on face adaptation. In face adaptation, exposure to a facial attribute (identity, expression, face attractiveness, etc.) biases the perception of subsequently presented faces in the direction opposite to that of the adapted attribute. To address the connection of the two aftereffects, we have previously shown that exposure to a concave curve biased the perception of subsequently presented faces’ facial expression judgment – the faces appear happier than they were originally, the cross-level adaptation. In the following studies, we aimed to reveal the mechanisms of face adaptation. We first investigated whether face aftereffect is generated de novo or inherited from the low-level features, and how does our visual system integrate facial emotion information from partial faces, or face streams presented in rapid serial visual presentation? We suggest that ensemble coding may be the mechanism behind this fast processing of the gist of facial emotion or face attractiveness information in the face stream. In a separate line of research, we investigate how we can change the environment to fit the needs of the human users for transport in virtual reality and real world environments.