Abstract
Ultra-rapid categorization studies analyze human responses to briefly flashed, static natural scenes. Recently, we reported that reaction times can be extremely fast if subjects are asked to move their eyes to the side where an animal had appeared. Accuracy was remarkably good with the fastest reliable saccades occurring in only 120 ms after stimulus onset [Kirchner & Thorpe, Vision Res, (2005)].
In the present study we determined the processing speed of static vs. dynamic displays. In blocked conditions, human subjects were asked to detect either animal or machine targets. On each trial, an examplar of each image category was presented simultaneously on the left and right of fixation for 80 ms. In half of the trials both images were static and in the other half of the trials both were dynamic (i.e., four consecutive images). Subjects made a saccade towards (or button press on) the side containing the target. While both response modes resulted in good accuracy, only saccades showed an advantage of dynamic over static trials. The saccade latency distributions indicated a clear dissociation in that dynamic displays resulted in much more continuous information accrual and a broad distribution as compared to a bimodal distribution in the static trials. A control study with simple form stimuli (squares vs. circles) served to preclude a low-level explanation of the result. We conclude that form processing can be improved by stimulus motion, and furthermore, that this information can be directly used to control directed behavior.
Supported by EU project “Perception for Recognition and Action” (HPRN-CT-2002-00226).