Abstract
A fundamental challenge in natural scene perception is the selection of reliable visual features to guide our daily navigation. Developmental and animal researchers have found that only certain cues facilitate this process. For example, 4-yr-olds reorient in accord with the geometry of a layout defined by walls only 30 cm high, but fail to do so when the walls are replaced by tape on the floor (Lee & Spelke, 2008). These studies suggest that three-dimensional surfaces serve as informative cues to define a functional space. The present research tests the relationship between different boundary cues and representation of the size of space. We used artificially-created indoor scenes that varied along 1) size of space (Small, Medium and Large), and 2) boundary cue (Wall, Curb and Line). Specifically, the boundary was defined by walls (Wall), vertical structure that was ~4 inches high (Curb), or flat lines on the floor (Line). Participants (N=4) viewed these scenes in blocks of eight seconds while performing a repetition task. We measured multivoxel pattern activity across scene-selective regions: parahippocampal place area (PPA), restrosplenial cortex (RSC), and transverse occipital sulcus (TOS). Using a linear classifier, we found a marginally significant classification in PPA, RSC and TOS (all ps <.1) for the different sizes, but not in LOC and FFA. Interestingly, these scene-selective ROIs did not show classification of the boundary cues. To test whether the extension of three-dimensional surface is important for defining the size of space, we compared the size classification for scenes defined by the Line to scenes defined by the Curb. RSC showed a significant classification for the size of space when defined by the Curb, but not when defined by the Line. These results are consistent with previous findings that show specificity for size of space property in RSC (Park et al., 2011).
Meeting abstract presented at VSS 2013