The perception of absolute speed during self-motion

Perceived velocity and smooth-pursuit eye movements for drifting gratings were shown to involve different combinations of signals (Gegenfurtner and Hawken, 1995, ARVO). We considered whether the estimation of absolute speeds of optic flow might also be different for a speed-matching task done by walking compared with a perceptual rating task. In the motor task, subjects were required to walk on a non-motorised, manually driven treadmill at their chosen pace while being presented with optic flow in the form of bright rectangles expanding against a dark background. In each trial an optic flow pattern was displayed for 10 s on a large rear-projected screen, and subjects were asked to adjust their walking speed to match the optic flow. Thirteen levels of optic flow ranging from 1 to 7 km h-1 were tested. In the rating task, standing subjects were required to make verbal judgments of speed (in km h-1) for the same optic flow presentations. We ran a third condition in which subjects rated as well as matched the flow by walking in each trial. The results for all subjects and conditions showed walking and rating speeds linearly related to the speed of the optic flow ( R2 ࣙ 0.91). However, the slopes of the regression lines were consistently lower for matched walking compared with the rating condition. Results from the third condition showed no direct effect of walking on the rating task. We conclude that visual velocity while walking relies on different mechanisms than those used for speed judgments based on visual signals alone