Browse by author
Lookup NU author(s): Dr Kun Guo, Dr Angel Nevado, Dr Robert Robertson, Maribel Pulgarin Montoya, Professor Alexander Thiele, Professor Malcolm Young
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
Spatial and temporal regularities commonly exist in natural visual scenes. The knowledge of the probability structure of these regularities is likely to be informative for an efficient visual system. Here we explored how manipulating the spatio-temporal prior probability of stimuli affects human orientation perception. Stimulus sequences comprised four collinear bars (predictors) which appeared successively towards the foveal region, followed by a target bar with the same or different orientation. Subjects' orientation perception of the foveal target was biased towards the orientation of the predictors when presented in a highly ordered and predictable sequence. The discrimination thresholds were significantly elevated in proportion to increasing prior probabilities of the predictors. Breaking this sequence, by randomising presentation order or presentation duration, decreased the thresholds. These psychophysical observations are consistent with a Bayesian model, suggesting that a predictable spatio-temporal stimulus structure and an increased probability of collinear trials are associated with the increasing prior expectation of collinear events. Our results suggest that statistical spatio-temporal stimulus regularities are effectively integrated by human visual cortex over a range of spatial and temporal positions, thereby systematically affecting perception. © 2004 Elsevier Ltd. All rights reserved.
Author(s): Guo K, Nevado A, Robertson RG, Pulgarin M, Thiele A, Young MP
Publication type: Article
Publication status: Published
Journal: Vision Research
Year: 2004
Volume: 44
Issue: 20
Pages: 2349-2358
ISSN (print): 0042-6989
ISSN (electronic): 1878-5646
Publisher: Pergamon
URL: http://dx.doi.org/10.1016/j.visres.2004.04.014
DOI: 10.1016/j.visres.2004.04.014
PubMed id: 15246751
Altmetrics provided by Altmetric
