Abstract
Visual crowding, the inability to recognize objects in clutter, sets a fundamental limit on visual perception and object recognition. It is assumed that, in the normal fovea, crowding does not exist or it only occurs over very small distances. Recently we showed that spatial and temporal crowding are correlated in the amblyopic fovea and normal periphery, suggesting a tradeoff between spatial and temporal processing of crowding. We showed (VSS 2013) that limiting stimulus availability using backward masking results in increased crowding and slower reaction time. Here we tested the crowding effect without backward masking. We used a method originally termed "contour interactions" to measure the surround effect on recognition of a single E letter size and target-flanker spacing of either 1 or 0.4 letter size, which is assumed to be at the border of the critical window of crowding in terms of letter spacing. We measured the crowding effect as a function of presentation time (30-240 msec). We also explored how a training consisting of detecting Gabor stimuli under spatial and temporal masking affects crowding (Polat et al., 2012). The training was carried out on iDevices at a distance of 40 cm (using the GlassesOff application). We found that robust crowding exists in the fovea: participants overcome crowding for long presentation times of 240 msec but exhibit crowding for short presentation times of 30 and 60 msec. After training, the crowding effect is reduced significantly, even for a presentation time of 30 msec. Thus, normal processing of foveal information is efficient for overcoming crowding with presentation times longer than 240 msec. Training improves processing speed, an effect that enables more efficient foveal processing to overcome crowding even for shorter presentation times. Our results suggest that temporal processing is a very important factor in processing of the crowding effect.
Meeting abstract presented at VSS 2014