Abstract
There is evidence that the visual system uses earth gravity (1g = 9.81 m/s²) in a prior-like fashion for different tasks. Relatively soon, we might, however, be exposed visually to gravities other than 1g (f. e. in augmented/ virtual reality). With this motivation, we designed an experiment to establish the human ability to distinguish between different visually presented gravities. To this end, we employed a two-interval forced choice paradigm: participants were shown spheres of tennis ball size approaching them on parabolic trajectories. These trajectories were governed by seven different gravities (from 0.7 g to 1.3 g in steps of 0.1 g), two horizontal velocities (6.0 m/s and 8.33 m/s) and two initial vertical velocities (3.7 m/s and 5.2 m/s). Each trial consisted of a pair of parabolas, one of which was governed by earth gravity and the other by one of the seven test gravities. The kinematic values were chosen such that trivial cues like height or presentation time did not by themselves predict the underlying gravity value unequivocally (f. e. the higher parabola was not always the one with lower gravity). Participants were asked to judge which of the two parabolas had the higher underlying gravity. Not all subjects could discriminate different gravities. For those who did perform above chance level, the response pattern was best explained by the temporal derivative of the elevation angle at the top point of the parabola, which depends on the value of gravity. This information may be used in a relative way by combining the rate of change of the elevation angle with the elevation angle itself. We conclude that, while the visual system is principally capable of distinguishing gravities, it does so in an indirect fashion by processing optic flow information dependent on the gravity value.
Meeting abstract presented at VSS 2017