hide
Free keywords:
-
Abstract:
There are large amounts of motion in scenes but only some are relevant to an observer, as these may indicate objects of interest (eg prey or predator). Here we used a visual-search paradigm to investigate the extent to which such motion plays a role in the perception of natural scenes.
Observers were presented with circular arrays of 2, 4, 6, or 8 static or dynamic scenes containing either human, animal, or machine movements. In experiment 1, observers searched for human targets among machine distractors on one block and machine targets among human distractors
on another block. The order of the blocks was counterbalanced. In experiment 2, we used humans and animals in the same design. In both experiments, half the observers were tested with static scenes and the other half were tested with dynamic scenes (N 12 per condition).
We found that observers were faster at searching for dynamic targets than for static targets. We also found category effects on search times. For both static and dynamic conditions, observers were faster at searching for humans among machines than they were at searching for machines among humans. However, there was no difference in search times between human and animal targets for dynamic conditions but there was a difference in search times for static conditions. Overall, the present results point to the importance of high-level interpretations of motion (eg as biological versus mechanical motion) in processing natural scenes.
