Abstract
When dissimilar images are presented to corresponding regions of the two eyes, the result is a series of perceptual switches between the two images. The visual system can switch between complete monocular images or those distributed piecemeal between the two eyes. Here it is shown that surface shape, defined by binocular disparity information, is sufficient to bias one of the rivalling images into dominance. Disparity was used to present either a convex or concave half-sphere with random dot texture. In the center of the hemisphere was a rivalrous circular region. Within this patch, monocular shape-from-shading information indicated that one eye's half-image was a convex bump and the other a concave dimple. Overlaid on each patch was a series of black bars; vertically oriented in one eye and horizontally in the other. During a 75-second trial, observers dynamically reported the orientation of the dominant bars. The results indicate that global shape does influence perception. When binocular disparity defined a convex shape, the monocular convex bump that completed the hemisphere was more dominant. When disparity defined a concave shape, the monocular concave dimple was more dominant. The results show that in cases involving multistable phenomena, the visual system can incorporate relevant information given by different visual cues in order to achieve a more coherent percept of the world.
Supported by the Wellcome Trust (UK) & NSF (USA)