Abstract
Previous research reported that both optic flow and target egocentric direction are used for the visual guidance of walking toward a goal. However, the effect of optic flow might be confounded with perspective cues provided by the structure of the environment. Here we disentangled the contribution of optic flow and perspective cues to the control of walking toward a goal. Participants (n=12) wore a head-mounted display (Oculus DK2, FOV: 100°) and walked toward a target placed in four types of virtual environment: (1) a vertical line that provided no optic flow or perspective information, (2) an empty room that provided perspective but no optic flow information, (3) a textured ground that provided optic flow but no perspective information, and (4) a textured room that provided both optic flow and perspective information. Participants' walking speed specified by optic flow was 0.5, 1, or 4 times (flow gains) of their physical walking speed. The target was placed in front of participants at 4, 8, or 32m in the virtual world according to the specified flow gain, and participants' virtual heading specified by optic flow was displaced ±10° from their physical walking direction. We found that overall path was straighter and heading error was smaller with than without optic flow. Perspective cues alone affected path curvature and heading error mainly at close distances. Perspective cues had minimal effects on walking in the presence of optic flow. With optic flow, path started to straighten and heading error started to drop earlier with increasing optic flow gain. We conclude that while perspective cues alone contribute to the control of walking at close distances, optic flow is a more powerful cue and diminishes the use of perspective cues. The larger the magnitude of optic flow, the sooner it is used for locomotion control.
Meeting abstract presented at VSS 2017