Purchase this article with an account.
Sheng He, Fang Fang; Integration of depth cues across space and modalities. Journal of Vision 2003;3(12):8. doi: 10.1167/3.12.8.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
3D structures can be perceived based on the patterns of 2D motion signals. This so-called structure from motion can sometimes result in bi-stable interpretations of the 3D structure. In one particular case, a 2D projection of a dotted rotating cylinder is perceived as a 3D rotating cylinder, but its direction of rotation is bi-stable, as reversals in the perceived direction of rotation occur involuntarily every few seconds. This is because the assignment of the front and back surface to the two sets of oppositely moving dots is ambiguous. Adding disparity information to the dots eliminates the ambiguity of the depth relationship thus stabilizing the bi-stable motion (Nawrot & Blake, 1989). We show that bi-stable motion can also be made unambiguous by adding information not to the dots themselves, but to the visual context. Furthermore, the stabilized bi-stable motion can generate consistent rotational motion aftereffects. The observed rotation aftereffect is disparity contingent, and is not caused by attentional tracking.
Information from a different modality can also help to disambiguate the visually bi-stable rotating cylinder. When haptic information is available from a rotating physical cylinder in the same location, observers tend to see the visually bi-stable cylinder rotating in the same direction. This tactile input to visual processing can be strengthened through training.
NawrotM.BlakeR.(1989). Neural integration of information specifying structure from stereopsis and motion. Science 244, 716–718.
This PDF is available to Subscribers Only