Introduction

The visual system tends to assign contrast borders to one of the adjacent regions as if they were contours of objects in 3D space. Without proper 3D information this assignment is generally ambiguous, as demonstrated by Rubin's vase figure (see left). Zhou, Friedman & von der Heydt (2000) have shown that many neurons in areas V1, V2, and V4 respond differently to contrast defined borders when these borders are displayed as part of visual objects on different sides (see Figure side selectivity), where the edge in the receptive field (ellipse) is the bottom left side of a square in A and C, but the top right side of a square in B and D. The cell shown here prefers the figure to be on the lower left of the RF irrespective of the contrast polarity. These findings suggested that neurons in these regions encode perceptual border ownership. The crucial test of this hypothesis is to examine the responses of orientation-selective V2 cells to contrast-defined and disparity-defined figures. A contrast-defined square is generally perceived as "figure", with the borders assigned to the square (see Figure side selectivity) while a corresponding region in a random-dot stereogram is perceived either as a "figure", if its disparity is "near", or as a "window", if its disparity is "far", relative to that of the surrounding region. In the stereogram, the nearer surface always owns the border. In this study we examine whether the figure side preference of neurons depends on disparity in the same way (see Stereograms). Cells that signal edges in random-dot stereograms exist in V2 (von der Heydt, Zhou & Friedman, 2000) (see Stereoscopic edge selectivity). If our hypothesis is correct, the disparity cue should override other figure ground cues such as convexity and size.