Occipital network for figure/ground organization

Abstract

This is the accepted version of the paper.This version of the publication may differ from the final published version. Abstract To study the cortical mechanism of Wgure/ ground categorization in the human brain, we employed fMRI and the temporal-asynchrony paradigm. This paradigm is able to eliminate any diVerential activation for local stimulus features, and thus to identify only global perceptual interactions. Strong segmentation of the image into diVerent spatial conWgurations was generated solely from temporal asynchronies between zones of homogeneous dynamic noise. The Wgure/ground conWguration was a single geometric Wgure enclosed in a larger surround region. In a control condition, the Wgure/ground organization was eliminated by segmenting the noise Weld into many identical temporal-asynchrony stripes. The manipulation of the type of perceptual organization triggered dramatic reorganization in the cortical activation pattern. The Wgure/ground conWguration generated suppression of the ground representation (limited to early retinotopic visual cortex, V1 and V2) and strong activation in the motion complex hMT+/ V5+; conversely, both responses were abolished when the Wgure/ground organization was eliminated. These results suggest that Wgure/ground processing is mediated by topdown suppression of the ground representation in the earliest visual areas V1/V2 through a signal arising in the motion complex. We propose a model of a recurrent cortical architecture incorporating suppressive feedback that operates in a topographic manner, forming a Wgure/ground categorization network distinct from that for "pure" scene segmentation and thus underlying the perceptual organization of dynamic scenes into cognitively relevant components. Permanent repository link

Neural evidence for competition-mediated suppression in the perception of a single object

“…The use of lateralized presentations uncovered a previously unobserved hemispheric asymmetry: competition-mediated ground suppression was only evident in the LH with RVF presentation, not in the RH with LVF presentation. Hemispheric differences were not observed previously because stimuli were either presented centrally (e.g., Likova & Tyler, 2008;Peterson & Kim, 2001;Peterson & Skow, 2008;Salvagio et al, 2012;Strother et al, 2012) or were presented in the RVF only and only LH activation was assessed (Kaster & Beck, 2005;Scalf et al, 2011). Although laterality effects are not uncommon in cognitive neuroscience, this result still raises the question of why our effect was only observed the LH.…”

“…One set of theories and models posit that two adjacent regions compete for object status (Craft et al, 2007;Grossberg, 1994;Kienker et al, 1986;Sejnowski & Hinton, 1987) with the winner perceived as the object, and the loser perceived as a shapeless ground. There has been some debate regarding whether neural responses are facilitated for perceived objects (Appelbaum, Wade, Vildavski, Pettet, & Norcia, 2006;Appelbaum, Wade, Pettet, Vildavski, & Norcia, 2008;Lamme, 1995), suppressed for grounds (Likova & Tyler, 2008;Peterson & Skow, 2008;Salvagio et al, 2012;Tsotsos, Culhane, Kei Wai, Lai, Davis, & Nuflo, 1995), or whether both effects occur (Strother et al, 2012). Here, our use of retinotopic mapping and precise localization of the cortical representations of the figure and ground allowed us to provide supporting evidence for ground suppression as a mechanism behind object perception.…”

“…Thus, it is likely that evidence of competition-mediated ground suppression observed in V1 and V2 is due to feedback from higher levels where receptive fields are larger. (For similar arguments involving fMRI data, see Likova & Tyler, 2008;Strother et al, 2012. )…”

“…We expected that the differential competition for object status in the two types of silhouettes and therefore any differences in ground suppression would be high in V4. Third, although RFs in V2 and V1 are much smaller (1-2 o ; Bles, Schwarzbach, De Weerd, Goebel, & Jansma, 2006) and cannot encompass more than 1-2 parts of the familiar object suggested on the groundside of the high-competition silhouettes, others have observed evidence for ground suppression in V1/V2 (Likova & Tyler, 2008;Strother et al, 2012). These authors attributed their V1/V2 results to feedback, because the pattern of activity in V1/V2 mimicked that in higher-level regions where stimulus differences between their conditions were represented (MT and LO for Likova &Tyler, 2008, andStrother et al, 2012, respectively).…”

“…For instance, Likova and Tyler (2008) showed that neural activity is suppressed for regions perceived as grounds compared to regions without figure-ground structure (cf., Strother, Lavell, & Vilis, 2012). Here, we extend those previous studies by using two types of stimuli that are matched for figure-ground structure (and for low-level features -see Section 2.1.2) to investigate whether more suppression is applied to grounds that compete more for object status in virtue of suggesting a familiar configuration (high-competition silhouettes) than to grounds that compete less for object status (low-competition silhouettes).…”