Research Project
7046004
DESCRIPTION (provided by applicant): A major goal of research on human form vision is to understand the sequence of operations that transform local contour orientation information extracted in primary visual cortex (V1) into global information about objects such as faces. Previous work has shown that intermediate levels of form vision (V4) analyze ellipsoidal, radial, and other configurations in the image. Based on this, we have devised a novel class of synthetic faces described by 37 geometric parameters. Psychophysical experiments using 4-dimensional "face cubes" have already shown that discrimination thresholds are lowest near the mean face for each gender and that face discrimination is probably based on principal components analysis (PCA). The proposed research will investigate the role of principal components in identifying faces across views and size differences, and masking will explore the temporal dynamics of face processing. The hypothesis is that faces are first transformed to a common view and then analyzed via PCA. The role of head rotation from front to side, which conveys additional 3- dimensional shape information, will also be assessed. Brain imaging (fMRI) will be used to determine the role of cortical areas MT and FFA in processing static and rotating synthetic faces. The fMRI experiments will also include quantitative measurements of FFA responses to principal components using an adaptation paradigm. Additional collaborative studies will use synthetic faces to quantitatively assess pattern recognition limitations of amblyopes and prosopagnosics (pilot data provided). These studies will provide new insights into the global processes underlying face perception and the ways in which they may be clinically compromised
