Purchase this article with an account.
Jennifer F. Schumacher, Christina F. Quinn, Cheryl A. Olman; Parameter exploration of contextually modulated collinear Gabor patches. Journal of Vision 2010;10(7):1158. doi: https://doi.org/10.1167/10.7.1158.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Context is an important factor that often hinders or helps one’s performance in a visual task. In a contour detection task, Dakin and Baruch, 2009 found that collinear Gabor patches in a parallel context increased the detection threshold while an orthogonal context decreased the threshold. However, our exploration of a parameter space including eccentricity, spacing, and spatial frequency revealed that this contextual effect is dependent on how the visual scene is arranged. Psychophysical data were collected in a 2AFC task where the subject determined which side of the visual scene (left or right) contained a set of five collinear Gabor patches. The eccentricity at which the collinear Gabors were presented (1.4–4.2 degrees), the spacing between Gabors (0.6–1.8 degree), and spatial frequency of the Gabors (2–4 cpd) varied across blocks of trials, for a total of eighteen configurations. The target Gabors were embedded in a grid of identical Gabors; tolerance to orientation jitter was measured using the method of constant stimuli (fifty trials at each of six levels of orientation jitter). Three different contexts were implemented in the surrounding grid: control, parallel, and orthogonal, where the orientations of the flanking Gabors were drawn from a distribution of orientations randomly oriented, parallel, or orthogonal to the target Gabors, respectively. For all three conditions, eccentricity had the greatest effect on the contour detection thresholds. The strength of orientation-dependent contextual effects varied over the parameter space, but depended most strongly on spatial frequency. The results also suggest that collinear facilitation and orientation-dependent modulation operate on different spatial scales implying involvement of different neural mechanisms. Overall, these results provide a systematic analysis of contour detection within a specific parameter space and while supporting that context influences performance in visual tasks, the magnitude of the effect is dependent on the composition of the visual scene.
This PDF is available to Subscribers Only