Nonvisual information contributes to flow parsing during walking

Abstract

To perceive the movement of an independently moving object during self-motion, visual motion caused by object movement must be distinguished from effects of self-motion. This has been termed flow parsing. Flow parsing could be accomplished from optic flow alone. During walking, nonvisual information about self-motion could also potentially contribute. In this study, we measured the relative influence of visual and nonvisual information on perception of object motion and movement control during walking. In a virtual environment, observers walked toward a distant target along a textured ground plane. An independently moving dot appeared briefly during movement, 0.9 m above the ground, and observers judged whether the object was moving leftward or rightward. The horizontal velocity of the object was varied across trials and responses were used to compute a point-of-subjective-equality representing the object motion that is perceived to be stationary. To dissociate visual and nonvisual cues, we presented conditions where the direction of self-motion indicated by optic flow differed from the physical direction of walking by ±5°. In this situation, the relative motion of an object that is stationary with respect to optic flow would not correspond to a stationary object when compared to the physical direction of self-motion, and vice versa. The object motion that is perceived to be stationary can therefore be used to infer the relative influence of visual and nonvisual cues. We also analyzed walking trajectories to estimate the relative influence of visual and nonvisual cues on movement control. We found that nonvisual information strongly influenced both object motion judgments and walking trajectories, with less weighting for object motion (60% vs. 90%). We also found trial-to-trial correlations between biases in walking direction and judgments of object motion. The results suggest that an integrated perception of heading contributes to both movement control and flow parsing. (Meeting abstract presented at VSS 2015)