Abstract
Morphogenesis—or the origin of complex natural form—has long fascinated researchers from practically every branch of science. Despite dramatic advances in our understanding of how physical and biological processes generate complex shapes like crystals, lungs, or sunflowers, we know practically nothing about how the human mind perceives and understands such processes. Inspired by D'Arcy Thompson's famous work, some researchers suggest the visual system may represent complex shapes and shape transformations as analogous distortions of a perceptual coordinate system. We put this idea to the test. If observers can infer non-rigid transformations, they should be able to estimate how points on or near an object shift in space as a result of the transformation. We showed participants pairs of objects ('before' and 'after' a transformation). On each trial, a probe location was indicated by a dot on or near the 'before' object and participants placed a second dot "at the corresponding location" on or near the 'after' object. Sampling many probe locations allowed us to map out in spatial detail how perceived shape and space were affected by the transformations. We tested four 2D stimulus sets of (1) drawings by D'Arcy Thompson, (2) unfamiliar shapes transformed by global non-rigid transformations, or (3) by local non-rigid transformations (growing 'limbs'), and (4) familiar real-world organisms. Participants' responses were strikingly accurate and mutually consistent for this wide range of shapes and non-rigid transformations. A zero free-parameter model based on matching and interpolating/extrapolating the positions of high-salience contour features predicts the data surprisingly well, suggesting observers infer spatial correspondences relative to key landmarks. Our findings reveal the operation of previously unknown perceptual organization processes that make us remarkably adept at identifying correspondences across complex shape-transforming processes. We suggest this ability is invaluable for many tasks that involve 'making sense' of shape.
Meeting abstract presented at VSS 2016