| |
Abstract:
Abstract: Human viewers are able to extract three-dimensional
(3D) shape information from the relative 2D motion of an object's
parts, an ability referred to as structure-from-motion (SFM)
perception. SFM has been investigated using random dot displays
that are rotated in a manner that is consistent with the projection
from the surfaces of a 3D object. These displays produce a clear
perception of 3D shape and demonstrate that motion cues are
sufficient, by themselves, to create vivid shape representations.
Given the ventral-dorsal pathway ("what" vs. "where") distinction
within the primate visual system, it is unclear which visual areas
are involved in extracting shape information from motion cues.
Motion information is typically confined to the dorsal pathway and
information about object shape is typically associated with the
ventral pathway. We identified the brain regions involved by using
functional magnetic resonance imaging (fMRI) while subjects viewed
SFM stimuli with various types of motion stimuli serving as
controls. Our results demonstrate that, in addition to dorsal
motion areas (e.g., MT+), ventral visual areas are involved in the
perception of motion-defined 3D shape. These results demonstrate
that motion information is not confined to the dorsal pathway and
that the recruitment of brain areas is dependent upon the
perception engendered by motion cues.
|
