Abstract
Many artists spend thousand of hours analyzing object structures in order to render them accurately. Here we ask whether this extensive training affects visual processing, specifically the analysis of object structure. First, participants’ drawing accuracy was assessed by comparing geometrical properties of their drawing to those of the original picture (Carson, Millard, Quehl and Danckert, 2012). We then used two tasks to evaluate the subjects’ efficiency of integration of object structure: visual search to measure how this efficiency changed with eccentricity, and masking to examine the rate of temporal integration at the fovea. In the first experiment, objects were displayed around a central fixation dot and participants had to report whether a target, a possible object among impossible distractors or vice versa, was present. Fixation was controlled with an eye tracker. We varied the objects’ eccentricity as well as their number, and object size was scaled with eccentricity. Despite the correction for cortical magnification, search performance decreased with eccentricity. However, there was a smaller decrease with eccentricity for subjects with better drawing scores (r=-.89, p<.04), suggesting more efficient processing of object structure in the periphery for these subjects. In the second experiment, a single object was centrally presented, followed by a dynamic mask. We varied object-mask SOA and participants had to report whether the object was structurally possible or not. The critical SOA at which performance reached 75% correct decreased with their drawing scores (r=-.99, p<.001), suggesting that more skilled subjects needed less time to encode and categorize object structure at the fovea. Taken together, these results suggest that more skilled individuals may call on faster central processing and a better peripheral integration of object structure.
Meeting abstract presented at VSS 2013