Purchase this article with an account.
Aidan P. Murphy, Hiroshi Ban, Andrew E. Welchman; The integration of texture- and disparity-defined slant in the human brain. Journal of Vision 2012;12(9):1193. doi: https://doi.org/10.1167/12.9.1193.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The visual system exploits multiple depth cues to achieve an accurate interpretation of the 3D environment. Psychophysical studies demonstrate that the integration of texture and disparity signals is near optimal, reducing variance when estimating surface slant. However, the cortical circuits responsible remain largely unknown. Here we test for cortical responses that relate to the integration of texture and disparity cues to slant. We measured human psychophysical judgments and fMRI responses to stimuli containing texture and disparity-defined slants. Participants viewed planes in which disparity and texture cues were independently manipulated to indicate one of two angles of slant (+60°/-30° rotation about the horizontal axis). Stimuli contained either ‘single’ cues (‘texture only’ and ‘disparity only’- in which case the other cue signalled 0° slant), or combined cues (‘congruent texture and disparity’ or ‘incongruent texture and disparity’). Psychophysical tests revealed that observers were more sensitive to slant discrimination for congruent combined cue stimuli compared to individual cues, the quadratic summation of individual cues or incongruent cues. The fMRI data were analysed by training a support vector machine (SVM) classifier to predict the slant angle (+60°/-30°) of stimuli based on patterns of activation in functionally localized regions of visual cortex. SVM classification accuracies for stimuli containing congruent texture and disparity cues exceeded those predicted by quadratic summation of single cue accuracies (i.e. the minimum bound for integration) in higher regions of both ventral and dorsal cortex (LO, V3B/KO and V7). This improvement was specific to congruent combinations of cues: incongruent cues supported lower decoding accuracies, which may suggest the robust use of individual cues in cases of large cue conflicts. These results are consistent with recent evidence that indicates that disparity and motion parallax depth cues are integrated in area V3B/KO, suggesting this area is intricately involved in the integration of multiple depth cues.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only