Visual Perception and the Guidance of Locomotion without Vision to Previously Seen Targets

Abstract

Two experiments were performed to assess the accuracy and precision with which adults perceive absolute egocentric distances to visible targets and coordinate their actions with them when walking without vision. In experiment 1 subjects stood in a large open field and attempted to judge the midpoint of self-to-target distances of between 4 and 24 m. In experiment 2 both highly practiced and unpracticed subjects stood in the same open field, viewed the same targets, and attempted to walk to them without vision or other environmental feedback under three conditions designed to assess the effects on accuracy of time-based memory decay and of walking at an unusually rapid pace. In experiment 1 the visual judgments were quite accurate and showed no systematic constant error. The small variable errors were linearly related to target distance. In experiment 2 the briskly paced walks were accurate, showing no systematic constant error, and the small, variable errors were a linear function of target distance and averaged about 8% of the target distance. Unlike Thomson's (1983) findings, there was not an abrupt increase in variable error at around 9 m, and no significant time-based effects were observed. The results demonstrate the accuracy of people's visual perception of absolute egocentric distances out to 24 m under open field conditions. The accuracy of people's walking without vision to previously seen targets shows that efferent and proprioceptive information about locomotion is closely calibrated to visually perceived distance. Sensitivity to the correlation of optical flow with efferent/proprioceptive information while walking with vision may provide the basis for this calibration when walking without vision.

Self-Motion Perception During Locomotor Recalibration: More Than Meets the Eye.

“…That is, having visually previewed a target as much as 20 m away (Rieser, Ashmead, Talor, & Youngquist, 1990), a person can close his or her eyes and walk to it fairly accurately and without bias. Indeed, walking without visual feedback is one of a number of indirect measures of distance perception (including throwing without feedback) at which humans perform especially well (e.g., Thomson, 1983; see Loomis & Beall, 2004, for a review).…”

Optic flow dominates visual scene polarity in causing adaptive modification of locomotor trajectory

“…Both the GVR and NGVR The GVR and virtual rotating rooms were cylindrically shaped with a simulated diameter of 20 m and height of 4 m and a rotating speed of 30 degrees per second to match with the treadmill walking speed. Previous studies have shown rotation rates close to this to be effective in causing sensations of self-motion [30,36]. The subject's simulated eyepoint was displaced 1.8 m above, 7 m left and 1 m back from the center of the both of the room's floor as it rotated about an axis coincident with its center, creating the illusion that the subject was walking around the perimeter of the room to his or her left.…”

“…Our previous study had shown a significant difference in the strategic torso stabilization responses during locomotion control when viewing polarized scenes in comparison to viewing non-polarized scenes that were rotated about or translated along all the degrees of freedom [36]. Other studies have shown the importance of optic flow and that optic flow in a virtual scene, clearly influences locomotion trajectory [44].…”

Manipulation of visual information does not change the accuracy of distance estimation during a blindfolded walking task

“…While many studies have shown, similar to Thomson, that blindfolded humans are able to navigate relatively successfully towards a target, his suggestion of a time limiting component has not been replicated (Fukusima, Loomis, & Da Silva, 1997;Rieser, Ashmead, Talor, & Youngquist, 1990), with many authors finding that participants become less accurate in their estimation if the distance is increased (Corlett, Patla, & Williams, 1985;Fukusima et al, 1997;Glasauer, Amorim, Vitte, & Berthoz, 1994;Loomis, Da Silva, Fujita, & Fukusima, 1992;Mittelstaedt & Mittelstaedt, 2001;Rieser et al, 1990;Steenhuis & Goodale, 1988). Indeed, there is good evidence to suggest that other factors apart from time can contribute to successful non-visual distance estimation.…”