Abstract
When an observer moves through a scene, visual and non-visual cues can aid heading estimation. One cue, the optic flow field, strongly influences perceived heading. Another cue, otolith stimulation signaling linear acceleration, also affects perceived heading (Sibigtroth & Banks, 2001). It is not known how these two cues are combined to determine one's heading. Specifically, we wondered what cues determine the weight of the otolith signal. There are two hypotheses: 1) the otolith weight increases with the magnitude of the otolith signal, and 2) the otolith weight increases with the visually-specified speed. We presented optic flow displays simulating curvilinear translation (100–800 cm/s) across a ground plane. Gaze was tangent to the path. We presented two path curvatures and these gave rise to rotation rates of 1–12 deg/s. To stimulate the otoliths, we exploited the fact that when an observer is on a curved path with centripetal acceleration x, the resultant gravito-inertial force is applied to the head at an angle whose sine is x. We simulated this force by rotating observers about the roll axis. To determine the relative influence of the flow field and the otolith signals, we independently manipulated the centripetal acceleration specified by the optic flow field, and that specified by the otolith signal (or equivalently, the roll angle). In particular, for a given path curvature and visually-specified speed, the visual stimulus did not vary with roll. Observers indicated their perceived path at the end of a 9-sec motion sequence. The majority of observers gave the otolith signal higher weight as visually-specified speed increased, consistent with hypothesis 2.