Purchase this article with an account.
Ilmari Kurki, Miguel P. Eckstein; How the perceptual template expands across the visual field with learning: a classification image study.. Journal of Vision 2012;12(9):699. doi: https://doi.org/10.1167/12.9.699.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual learning changes the way the visual system samples and integrates the stimulus information. Efficiency of the internal weighting of the stimulus information (perceptual template) has been found to improve, explaining at least partially the performance improvement. However, less is known about the principles that determine how learning changes the template. Here we use classification image (CI) method to investigate how the tuning of observer's template change in line orientation discrimination task. Previous studies (Li, Levi & Klein, 2004; Dobres & Seitz, 2010) have found that observers initially use mostly foveal information, while as learning progresses the template spreads towards the periphery. However, the stimuli in these studies have also had the highest signal to noise ratio (SNR) in the fovea, leaving two potential theories for the change: 1) a spread from the fovea to the periphery; 2) a spread from the stimuli components with the highest SNR. Here, stimuli were noisy peripheral oriented lines where the SNR was inversely related to the distance from the center. Lines were composed of 16 windowed checkerboard elements (width 0.75°, spacing 1.0°) displayed at 10°. Eye tracking was used to ensure central fixation. Underlying orientation of the lines was varied using method of constant stimuli, and positional noise (Li, Levi & Klein, 2004) was added to perturb the element locations. Observers discriminated the line orientation using a rating scale yes-no task (800 trials / session, 5 sessions). All observers showed steady improvement in the discrimination performance. CIs show that initially observers used a sub-optimal strategy, relying almost solely on the top and bottom parts of the stimulus where the SNR was highest. During the learning observers tuned their templates, extending the weighting especially to adjacent locations close to the initial hotspot. Template change was most pronounced in observers having the worst initial performance.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only