Abstract
Holcombe and Cavanagh (2001) introduced a psychophysical method that enabled them to place an upper bound on the speed at which the human visual system can integrate color and orientation. They concluded that the visual system requires, at most, 25 milliseconds to determine how color and orientation are conjoined. However this method does not quantify the temporal resolution of color and of orientation, and thus it does not place a lower bound on the time required for feature integration. Indeed, if 25 milliseconds were required to process either feature, this would rule out a time-consuming integration computation. Here we introduce a psychophysical method that provides comparable measures of temporal resolution for color, orientation and color/orientation conjunctions. Since all measures are derived based on judgments of the same stimuli, the method is robust with respect to changes in arousal and adaptation state. Preliminary results from three subjects suggest that the temporal resolution for color/orientation conjunctions is slower than temporal resolution of either feature, providing evidence for a non-instantaneous integration computation.
Funding provided by NEI grant R01-EY13802-01