Abstract
The ambient visual information that contributes to self- and object-orientation includes frame, horizon, and visual polarity cues (derived from sources such as the direction of illumination, the relationship between objects, and intrinsic polarity cues). These cues can be ambiguous: the frame provides four possible directions of up; the horizon two; whereas polarity cues provide a unique up direction. We previously showed how these elements affect the perceptual upright using oriented gratings (http://journalofvision.org/5/8/193). Here we look at the relative contribution of each component in real world scenes.
Using the Immersive Visual Environment at York (IVY) we placed eleven observers in a virtual reality simulation of (i) a fully furnished room, (ii) just the furniture from this room, (iii) the room without furniture, (iv) a “room” comprised of random dots and (v) a wire-frame room. The environments were rendered in stereo using shutter glasses and could be tilted relative to the viewer. For each orientation of each environment, upright observers performed the Oriented Character Recognition Test (Exp Brain Res. 173: 612) to estimate the perceptual upright (PU). The PU was modeled as the sum of four vectors corresponding to the body/gravity (one vector) and the three visual components (frame, horizon, and polarized cues).
The contribution of the wire-frame and dot rooms was dominated by the frame. Interestingly the relative contribution of the frame was similar when all the other cues were present (15% wireframe, 11% empty room, 19% furniture-filled room) but also when the frame was only implied in the furniture-only display (13%). The effect was largest for the furniture-filled room (83% of the body+gravity cue, furniture alone 74%, empty room 47%, wire-frame 14%, random dot 5%).
The different components of a visual scene make differential contributions to the perceptual upright that can be quantified precisely.
Supported by NASA Cooperative Agreement NCC9-58 with the National Space Bio-medical Research Institute, the Canadian Space Agency, and grants from the Natural Sciences and Engineering Research Council of Canada and the Canadian Space Agency (CSA).