|
Abstract:
Selective attention allows us to attend to particular parts
of a visual scene. We investigated whether patients with unilateral
hemispatial neglect are impaired in their ability to allocate their
attention to locations in the scene (location-based attention) or
in their ability to allocate attention to an object (object-based
attention). Tipper and Behrmann (1994) have argued for an
object-based frame of reference in object-based neglect. Although
Mozer (1998) has shown that a computational neural network model
(MORSEL) can account for object-based neglect by coding for the
locations of features in a location-based frame of reference. The
current study investigates hypotheses generated by the model in
neglect patients using a paradigm modified from the Tipper and
Behrmann (1994) study. Results support the model's attentional
mechanism. "Object-based" neglect was found using a partially
connected display that consisted of two objects. These results
suggest the relative location of features in an initial
location-based frame of reference are critical in the object
segmentation of visual scenes.
|
