Akshita Reddy Mavurapu¹ (), Manish Singh², Ömer Dağlar Tanrikulu¹; ¹University of New Hampshire, ²Rutgers University, New Brunswick

Humans can report the physical stability of objects with remarkable accuracy, highlighting the predictive power of visual perception (Cholewiak et al., 2013). In this study, we investigated whether physical stability influences figure-ground segmentation by manipulating the relative physical stability of two sets of regions. We created black-and-white repeated-region figure-ground stimuli to investigate how relative physical stability influences figure-ground assignment. The relative stability of adjacent regions was manipulated by using sine waves as boundaries, with variations in the sine waves' amplitude altering the height of the center of mass across the regions, thereby affecting their stability. The regions were carefully designed to eliminate biases from other well-known figure-ground cues, such as symmetry, convexity, and area while maintaining equal widths at the top and bottom to ensure that top-bottom polarity remained constant across stimuli (Hulleman & Humphreys, 2004). Participants viewed each stimulus for 300 ms and indicated whether they perceived the black or white region as the figure. Results showed that participants were more likely to choose regions with a lower center of mass as the figure. By varying the center of mass, we also indirectly altered the critical angle, which observers are known to be sensitive to when assessing an object's stability. The results suggest that stability can influence figure-ground assignment, even when controlling top-bottom polarity. The findings demonstrate that stability influences figure-ground perception, revealing a potential visual processing bias favoring stable objects. While these results introduce stability as a novel cue to figure-ground organization, we propose that stability may represent a more fundamental bias underlying cues previously reported in the literature, such as top-bottom polarity. These cues could be interpreted as exceptional cases of a broader stability-based bias in visual processing.