Item

Abstract

I investigate how humans solve the underconstrained problem of spatial extrapolation of complex 2-D curves. Contour continuation is typically modeled using local variational principles such as minimizing total curvature or variation of curvature and have only been applied to a very limited set of curves classes. Singh and Fulvio (2006) showed that only a probabilistic model of these constraints can explain the continuation of arcs of Euler spirals. I investigate whether this model generalizes to more complex curves such as polynomials or regular curves, how much higher-order information is used, and how the extrapolation is influenced by the experimental design. Subjects report either the position of a single curve dot behind a semi-circular occluder or draw the distribution of possible extrapolants. Results show that subjects report multiple extrapolants for more complex curves if they are given the opportunity (by drawing multimodal distributions or by probability matching). I model the results with Bayesia
n model averaging, and show that this explains the data better than a static model. This suggests that observers do not use a fixed extrapolation scheme but can also adapt to the context, taking into account the complexity of the visible portion of the curve.