Integration times of global shape mechanisms are limited by low level processes operating prior to spatial integration
53.4047, Tuesday, 19-May, 8:30 am - 12:30 pm, Pavilion
Jason Bell1, Gideon Sacks1, David Burr1,2; 1School of Psychology, University of Western Australia, Australia, 2Department of Neuroscience, Psychology, Pharmacology and Child Health, University of Florence, Italy
The temporal integration times of mechanisms detecting sinusoidal gratings, likely in V1, vary considerably, depending on spatial scale (BURR Proc Roy Soc, 1981). Here we investigate whether the temporal properties of higher level form mechanisms similarly depend on spatial scale, as would be expected if the temporal limitation were low level. We measure spatial and temporal summation in mechanisms responsible for decoding visual shape, using Radial Frequency (RF) patterns, globally processed stimuli that represent familiar shapes and objects, while being easy to control. We measured pattern deformation thresholds for one and three cycles of an RF3 (triangular pattern), as a function of SF profile (1.25 or 10 cpd), and presentation time (10-320ms). In all conditions, stimuli were matched for perceived contrast. SF and contrast matched backward masks were used to disrupt visible persistence. The results show that sensitivity to RF shape improved sharply as presentation time increased. The improvement was more rapid, and quicker to plateau for the low, compared with the high SF profiled pattern; indicating a shorter temporal integration time for the former (consistent with results with simple sinusoidal patterns). Sensitivity also improved with number of features, almost linearly (as previous shown: LOFFLER et al., Vis Res, 2003), irrespective of SF or presentation time. That the temporal properties of shape mechanisms vary with SF, in a similar way as for simple sinusoids, suggests that this integration may be limited at low levels, before spatial integration. This is consistent with the fact that there is not interaction between spatial integration and exposure time, as may be expected if the limitation were at a higher level. We will discuss these results in relation to the temporal properties of other visual mechanisms.