Search

The effect of visual experience is usually investigated through active (task dependent) training in a discrimination task. In contrast, the current work explored the psychophysical and electrophysiological correlates of passive (task independent) visual experience in texture segmentation by using an inattentional blindness-like paradigm (Mack et al., 1992). The psychophysical and electrophysiological responses to a segmented line-texture bar, with texture elements oriented either congruently (parallel) or noncongruently (orthogonal) to bar orientation, were collected after both short and long passive experience, with the texture presented on the background while subjects performed a primary task. Subjects were not able to distinguish the orientation of the bar (psychophysical results) after either short or long passive experience. However, the short experience produced an electrophysiological correlate of texture segmentation (N150), and the amplitude of this component was greater for the parallel bar, demonstrating that it reflected not simply local orientation discontinuities but also texture boundary-surface orientation congruency. This configurational effect in texture segmentation, which occurred without awareness during passive viewing, disappeared when the subjects had previously discriminated the orientation of the bar and when experience was lengthened, probably as a consequence of adaptation. Our study provides the first ERP evidence that boundary-surface relations are available during short passive visual experiences of very salient texture images and are suppressed by long experience, probably because of adaptation., (Copyright © 2011 Elsevier Ltd. All rights reserved.)

Gestalt factors of collinearity and similarity facilitate two fundamental perceptual tasks: grouping elements into figures and segmentation of figures from the ground. We have used a global-local paradigm to examine the psychophysical and neural correlates of these processes in humans: observers discriminated between orientations of either a three-Gabor group (grouping), or of a central Gabor within the group (segmentation). Groups were centered on a background of differently oriented Gabors. In both tasks, accuracy was increased by the collinearity (Experiment 1) and similarity (Experiment 2) of elements within the three-Gabor group. ERP correlates of facilitation differed across tasks. For segmentation, they were indexed by increased amplitude of negative ERP components, specific for processing textures, peaking at 75-250 and 150-250 ms, respectively. For grouping, collinearity and similarity had different effects. Collinearity produced a positive polarity deflection between 40 and 179 ms (i.e. the opposite to segmentation). This task-dependent switch in sign of polarity change, without corresponding changes in the stimulus or perception, reflects distinct neural mechanisms for collinear facilitation in grouping and segmentation. In contrast, similarity reduced positivity at 275 ms. Results show similar modulation of segmentation components via the distinct mechanism underlying collinearity and similarity, but distinct modulation of grouping components via collinearity and similarity.