Selective Visual Perception
Our research suggests that vision uses a pre-attentive (i.e.
consciously inaccessible) mechanism
that can detect an important visual pattern, as defined by a broad
range of criteria, from low level features up to
semantic meaning. Detection triggers
the deployment of a transient attentional burst rapidly enough to occur
within the same visual fixation (i.e. about 200 milliseconds).
This attentional burst is intended to snatch the target item from the
visual stream and encode it into memory before the eye moves away.
Most recently, this research has studied the ability of a target
item to capture attention (Wyble, Potter & Bowman under review).
My research involves a combination of methods: Behavioral
data
from paradigms such as rapid serial visual presentation (in which we observe the Attentional Blink
and Repetition Blindness) find the limits of the visual system's functional
capabilities. From these data, neurally inspired computational models provide insight into the sequence of the cognitive
operations involved in perception.
As an additional source of convergent evidence, electrophysiological
data can be compared to the time course of activity within the model.
Human EEG is particularly well suited to identifying the
temporal
correlates of selection
and encoding, thought to be
indexed by the N2pc and P3 components, respectively.
In this vein, Patrick Craston and Srivas Chennu, both grad
students at the University of Kent, have provided important theoretical
work and EEG data in relation to the aforementioned modelling efforts.