The goal of this research is to examine to what degree and for which experimental conditions attentional allocation is stimulus-driven in human observers. The proposed experiments will measure attentional allocation for natural stimuli and a range of experimental tasks thought to induce either stimulus-driven or goal-directed attentional allocation. The stimulus-dependence of attention will be measured by comparing the predictions of a purely stimulus-driven computational model of selective attention that is based on the neural mechanisms of the primate visual system with behavioral measures of attention. Eye movements will be recorded from participants as they view natural stimuli and used as a measure of attention. This research will extend our understanding of attentional allocation towards complex natural stimuli and clarify the effect of different behavioral tasks on the way in which attention is allocated.
