Skeletal descriptions of shape provide unique perceptual information for object recognition

Abstract

With seemingly little effort, humans can both identify an object across large changes in orientation and extend category membership to novel exemplars. Although researchers argue that object shape is crucial in these cases, there are open questions as to how shape is represented for object recognition. Here we tested whether the human visual system incorporates a three-dimensional skeletal descriptor of shape to determine an object’s identity. Skeletal models not only provide a compact description of an object’s global shape structure, but also provide a quantitative metric by which to compare the visual similarity between shapes. Our results showed that a model of skeletal similarity explained the greatest amount of variance in participants’ object dissimilarity judgments when compared with other computational models of visual similarity (Experiment 1). Moreover, parametric changes to an object’s skeleton led to proportional changes in perceived similarity, even when controlling for another model of structure (Experiment 2). Importantly, participants preferentially categorized objects by their skeletons across changes to local shape contours and non-accidental properties (Experiment 3). Our findings highlight the importance of skeletal structure in vision, not only as a shape descriptor, but also as a diagnostic cue of object identity.

Can humans perform mental regression on a graph? Accuracy and bias in the perception of scatterplots

“…Interestingly, object perception is not limited to the perception of a single straight axis (the principal axis), but can accommodate the perception of complex objects or spaces comprising multiple, possibly flexible subparts. In the case, the visual system appears to extract multiple axes of quasi-symmetry that may correspond to the mathematical concept of "medial axis" or "shape skeleton" (Ayzenberg & Lourenco, 2019;Cohen & Singh, 2006;Kelly & Durocher, 2011;Kovacs & Julesz, 1994;Lowet et al, 2018). Mathematically, a medial axis is the locus of points equidistant from the two closest shape boundaries (Blum, 1967(Blum, , 1973.…”

The contribution of environmental and dispersal filters on phylogenetic and taxonomic beta diversity patterns in Amazonian tree communities

“…The longstanding debate concerning the role of ecological interactions and environmental filters vs. neutral processes as mechanisms explaining tree community assembly in Amazon forests is a prominent topic in ecological and evolutionary research (Tuomisto et al 2002 ; ter Steege et al 2006 ; Kraft et al 2011 ; Swenson et al 2013 ; Baker et al 2014 ; Pos et al 2019 ). Amazonian forests are among the most diverse in the world harboring around 16,000 tree species with some local communities containing over 300 tree species in a single hectare (ter Steege et al 2013 ; Guevara et al 2019 ; ter Steege et al 2019 ; ter Steege et al 2020 ). Thus, describing and analysing the patterns of species and lineage composition across spatial and environmental gradients is fundamental to understand the mechanistic causes that promote community assembly of this hyper-diverse ecosystem.…”

Hiding the Rabbit: Using a genetic algorithm to investigate shape guidance in visual search

“…The number of iterations was set to infinity. We quantified the similarity between the shape skeletons of target and distractor shapes for each generation using the skeleton similarity metric from Ayzenberg and Lourenco (2019) . Briefly, this method calculates the mean Euclidean distance between each point on the distractor skeleton and the closest point on the target skeleton following maximal alignment.…”

“…For most shapes, the axes are organized in a hierarchical fashion such that a set of parent axes describe the coarse global geometry of the shape, while smaller axes describe individual and smaller component parts. Importantly, methods have been formalized to compare the similarity between shape skeletons, providing a quantitative metric to compare shapes by (Sebastian, Klein & Kimia, 2004 ; Ayzenberg & Lourenco (2019)) . A number of psychophysical results suggest a privileged role for skeleton representations in human object perception ( Kimia, 2003 ; Kovács, Fehér, Julesz, 1998 , Siddiqi et al, 2001 , Wilder et al, 2011 ).…”

“…In this case, the effects of evolution seem to be best captured by the shape skeleton measure from Ayzenberg and Lourenco (2019) . This may be intuitively clear from inspection of Figure 14 , where it seems that the hard distractors have a main axis and a part or two whereas the easy distractors lack that main axis.…”