Visual Attention Can Affect Perceived Distance

We can accurately perceive the egocentric distance of an object placed at a distance from0to20m in the light environment. In the dark where environmental cues are not visible, the perceived position of a lit target locates at the intersection between the projection line from the eye to the target and an implicit surface. The implicit surface, called intrinsic bias, can be approximated as a curved surface of which the slant increases with the egocentric distance. When the linear perspective texture (dim lit elements), a very effective depth cue, is placed horizontally in the dark, a lit target within the horizontal plane of the texture is judged on a less slant surface. Our visual system is more efficient in representing texture surface in the lower visual field (ground) than the upper visual field (sky/ceiling). Since the visual system uses the large background surface such as the ground surface as reference frame for space perception, perceived target locations is more accurate in the lower field (ground dominance). To further understand the mechanisms underlying space perception, we tested the hypothesis that visual attention plays a significant role in constructing background surface representation and thus space perception. We conducted two experiments in the dark with or without linear perspective textures.In Experiment1, we took a2(direction attention shifts to:same with textures, opposite to textures)×3(height of linear textures:0.35m,2*eye-height-0.35m, no textures)×3(target egocentric distances:4.5,6,7.5m)×2(target vertical heights:0.35m,2*eye-height-0.35m) within subject design and used the blind walking-gesture paradigm to measure the walked distance and pointed height of the target (thus target location). The perceived distance and height are combined to compute the perceived eye-to-target distance. The results showed that for the target in the lower field, the perceived eye-to-target distance was longer when attention and textures overlap (cue-valid) than when they did not (cue-invalid). But for the target in the upper field, the perceived eye-to-target distances were similar in the cue-valid and cue-invalid conditions. In addition, the targets in the lower field were perceived farther than that in the upper field when there were linear textures. Our results reveal that the use efficiency of the linear textures when attention and textures are in the same field is higher than that when they are in the different field, which reveals the important role of visual attention in space perception.In Experiment2, we presented the target at the eye level and the linear textures in the upper and lower field simultaneously. A2(directions attention shifts to:upper field, lower field)×2(textures background:present, absent)×4(target egocentric distances:3.25,4.5,6,7.5m) within subject design was used. We found that when the attention was shifted to the lower field, the perceived eye-to-target distance was longer than that when attention was shifted to the upper field. Since we only manipulated attention deployment during the experiment, this finding provides further supports for the attention hypothesis.