Contrast detection threshold of a Gabor signal (GS) is enhanced in the presence of high contrast GS flankers. Repetitive performance of the task induces contrast sensitivity changes on different time scales: minutes when using mental imagery, and weeks in the context of perceptual learning. We tested the time course of lateral enhancement within a single trial, using a forward masking technique. Contrast detection thresholds were measured (2AFC) for a foveal GS target presented briefly (32 ms) preceded by a presentation (80 ms) of two high-contrast GS flankers, with stimulus onset asynchrony (SOA) varying from 0 to 3600 ms. Using target-to-mask separation of 3\lambda and 12\lambda (\lambda =18°, GS wavelength), we found that the 3\lambda separation GS flankers decreased target threshold by 0.25 log units at SOA=0 and by 0.17 log units at 3600 ms. At 12\lambda separation, threshold decreased by 0.11 log units at SOA=0 and by 0.14 log units at 3600 ms. Long-term and short-term enhancements showed similar dependence on stimulus configuration (maximal for collinear target and masks) and local parameters (orientation and spatial frequency differences between target and flankers). The results imply that spatial filters in early vision retain an input trace for as long as a few seconds (up to 14.4 seconds tested). This trace may subserve the consolidation of filter activity into long-term memory.
Eccentricity dependence of orientation anisotropy of surround suppression of contrast-detection threshold
