Abstract
Previous studies suggest that the human perceptual system may be particularly tuned to biological movements, thus allowing for faster detection of biological (e.g., humans) as opposed to non-biological objects (e.g., cars) in the environment. This sensitivity to biological motion has been demonstrated for degraded stimuli, such as point light walkers, and for natural dynamic scenes. Using a visual search task, we previously showed a decrease in search time when observers searched for human targets compared to non-human targets. To gain further insights into how observers detect target scenes amongst distractor scenes, we combined response-time measurements with eye-movement measurements, as eye movements are thought to reflect search strategies that observers may use. Circular search arrays of two, four, six or eight movies were shown on a monitor. Participants searched for a target movie which was either present or absent in this array while their eye movements were tracked by an infra-red eye tracker. On two separate blocks, targets were either movies displaying humans or machines. Distractors were movies displaying the other category. For both human and machine targets, search times increased with set size. However, observers were faster in detecting human target movies amongst machine distractors than vice versa. This advantage was also shown in the number and duration of fixations to target scenes. Overall, these results suggest that movies displaying humans capture more attention than movies displaying machines, which is in accordance with the idea that the perceptual system is tuned to detect biological motion.
We would like to thank Yoav Tadmor for usage of facilities.