Abstract
Although research on human oculomotor control has found low sensitivity to acceleration, humans in natural environments have little trouble tracking a ball approaching in depth prior to an attempted catch. This is surprising because a ball moving along even a constant velocity through Euclidean space will undergo high values of angular acceleration as it moves towards a location near to the head. In this study, we test the hypothesis that humans compensate for the ball's angular acceleration by coupling movement of the gaze vector to the ball's pattern of optical expansion - a perceptual covariate of optical acceleration due to movement in depth. To test this hypothesis, subjects immersed in a virtual reality ball catching simulation were tasked with catching a ball that travelled to one of three locations at distances scaled to the subject's maximum reach. Arrival height was randomized within a range centered around head height. Although initial ball radius (cm) was constant, modified expansion rates were brought about by algorithmic manipulation of the ball radius through deflation/inflation during flight, consistent with the application of a gain term (δ) to the expansion rate. A gain of 1 is a consistent with the natural rate of expansion of an approaching constant sized ball, δ< 1 is diminished expansion due to deflation, and δ>1 is exaggerated expansion due to inflation. Subjects performed 10 repetitions at gain values of 0.5, 0.75, 1, 1.25, and 1.5 (3 passing distances x 5 gain x 10 repetitions = 150 trials per subject). The results indicate that the gaze vector shifted further along the ball's trajectory at greater values of δ, consistent with the hypothesis that subjects in the natural context account for angular acceleration by coupling movement of the gaze vector to the ball's rate of angular expansion.
Meeting abstract presented at VSS 2018