The influence of non-visual signals of walking on the perceived speed of optic flow

A E I Thurrell1, A Pelah1, H K Distler2

1 Department of Physiology, University of Cambridge, Downing Street, Cambridge CB2 3EG, UK
2 Max-Planck-Institut für biologische Kybernetik, Spemannstrasse 38, D 72076 Tübingen, Germany; fax: +44 122 333 3840; e-mail: ap114@cam.ac.uk

We considered how non-visual signals that accompany walking might influence the visual processing of optic flow. During natural locomotion optic-flow speed is determined by walking speed in a closed-loop manner. In the experiments subjects were required to adjust the speed of an optic-flow pattern to match that of a reference flow pattern in an open-loop manner. The visual speed was matched while walking by turning a hand-held knob that controlled the presented optic flow speed. Subjects were also required to change their pace according to a written instruction at the beginning of each trial to either 'very slow,' 'slow,' 'normal,' 'fast,' or 'very fast'. A non-motorized, self-driven treadmill simulated natural walking by allowing subjects to walk at their chosen pace. The optic-flow pattern consisted of bright rectangles expanding against a dark background displayed on a large rear-projected screen. An experimental block consisted of a 15 s presentation of a reference pattern followed by 5 test trials, one at each pace randomly ordered. Results showed a consistent increase of matched optic-flow speed with increasing walking speeds. However, speed settings were most accurate when subjects were walking at their chosen 'normal' walking pace. We conclude that the perceived speed of optic-flow information is influenced by non-visual signals (e.g. proprioception) such that optic flow speed is overestimated at lower walking speeds while underestimated at higher speeds.

Up    PDF (157kB)