Abstract
In natural behavior fixation patterns are tightly linked to the ongoing task. However, it is not known how task control interacts with image properties. How do task driven systems deal with unexpected salient stimuli? Previously, we studied the effect of potential collisions with pedestrians on gaze distribution of Ss walking in a virtual environment. Ss often failed to fixate pedestrians on a collision path, suggesting that potential collisions do not automatically attract a fixation, unless the observer is actively monitoring for pedestrians in peripheral vision. In this study we increased the saliency of colliders by increasing their speed by ∼25% during the collision period, compared to the previous condition where they only changed direction. About 10% of pedestrians on a non-colliding path changed onto a collision course for 1 sec, and then returned to the original, non-colliding path. If peripheral vision is constantly monitored, or the deviation attracts attention, this should be revealed by a fixation. In one condition Ss were instructed to follow a virtual pedestrian leader, and in another condition, to walk at their natural pace. Ss are most likely to fixate pedestrians in the first 2 secs after they appear, during which time, in non-leader trials, Ss fixate normal (non-colliding) pedestrians about 58% of the time; speeding colliders are fixated about 88%, and non-speeding ones about 68% of the time. However, this difference was not maintained for the remaining 3 secs that pedestrians are typically in the field of view. Thus speeding colliders have added power to attract attention only when observers are likely to fixate pedestrians in general. This interpretation is supported by the results in the leader condition where the probability of fixating normal pedestrians in the first 2 secs is reduced to 30%, to 37% (speeding), and to 31% (non-speeding colliders). Thus a competing task diminishes the ability of the speeding colliders to capture attention.
NIH grants EY05729 and RR09283