<title>Adaptive neural methods for multiplexing oriented edges</title>

Abstract

Edge linearization operators are often used in computer vision and in neural network models of vision to reconstruct noisy or incomplete edges. Such operators gather evidence for the presence of an edge at various orientations across all image locations and then choose the orientation that best fits the data at each point. One disadvantage of such methods is that they often function in a winner-take-all fashion: the presence of only a single orientation can be represented at any point; multiple edges cannot be represented where they intersect. For example, the neural Boundary Contour System of Grossberg and Mingolla implements a form of winner-take-all competition between orthogonal orientations at each spatial location, to promote sharpening of noisy, uncertain image data. But that competition may produce rivalry, oscillation, instability, or mutual suppression when intersecting edges (e.g., a cross) are present. This "cross problem" exists for all techniques, including Markov Random Fields, where a representation of a chosen favored orientation suppresses representations of alternate orientations.A new adaptive technique, using both an inhibitory learning rule and an excitatory learning rule, weakens inhibition between neurons representing poorly correlated orientations. It may reasonably be assumed that neurons coding dissimilar orientations are less likely to be coactivated than neurons coding similar orientations. Multiplexing by superposition is ordinarily generated: combinations of intersecting edges become represented by simultaneous activation of multiple neurons, each of which represents a single supported oriented edge. Unsupported or weakly supported orientations are suppressed. The cross problem is thereby solved.

The perception of globally coherent motion

“…9 overlapping contours. Several researchers have analyzed the special circumstances induced by line ends and comers in the domain of static form perception, (Grossberg & Mingolla, 1985a, 1985bMarshall, 1990a, Walters, 1987, Zucker, Dobbins, & Iverson, 1989, and in the motion domain (Grossberg & Mingolla, 1990;Marshall, 1990b.) It may be that measures of component velocity of the accuracy needed for the intersection-of-constraints construction arc indeed feasible for our visual systems for pure periodic stimuli, such as sinusoidal gratings, despite the overlap of different orientations.…”

“…(1985) involves an intersection-of-constraints solution in velocity space. (See Figure 1.) Recent experiments by Ferrera and Wilson (1987, 1990 have provided compelling evidence, however, that a velocity space solution cannot provide a complete account of observers' perceptions of moving patterns. For certain types of displays, which they refer to as Type II motion (as in Figure 2) observers seem to adopt a compromise between a velocity space solution based on intersection-of-constraints and a simple vector average of the individual component motions.…”

Models of receptive-field dynamics in visual cortex

“…However, isotropic RFs are not essential for these networks to produce changes in RF properties during the various forms of input conditioning. The EXIN and LISSOM learning rules are competitive learning rules and produce orientation selective neurons if the input features are oriented (Marshall, 1990d;Sirosh et al, 1996); scotoma conditioning using oriented features would a ect the networks as described in Sections 2.5.4 and 2.6.3. The adaptation networks are based on neuronal adaptation, without synaptic plasticity.…”

“…The EXIN a erent excitatory and lateral inhibitory synaptic plasticity rules together have been used to model visual disparity selectivity (Marshall, 1990c), visual motion selectivity and grouping (Marshall, 1990a), visual inertia (Hubbard & Marshall, 1994), visual motion integration in the aperture problem (Marshall, 1990a), visual length selectivity and end-stopping (Marshall, 1990b), visual depth perception from occlusion events (Marshall & Alley, 1993;Marshall et al, 1996a), visual depth from motion parallax (Marshall, 1989), visual motion smearing (Martin & Marshall, 1993), visual orientation selectivity (Marshall, 1990d), and visual stereomatching (Marshall et al, 1996b).…”

<title>Challenges of vision theory: self-organization of neural mechanisms for stable steering of object-grouping data in visual motion perception</title>

“…Self-organization methods, using unsupervised learning rules, have been useful in modeling the development of several aspects of low-level visual perception, including sensitivity to contrast (Linsker, 1986a), orientation (Bienenstock, Cooper, & Munro, 1982;Linsker, 1986b;Marshall, 1990f;von der Malsburg, 1973), motion (Coolen & Kuijk, 1989;Földiák, 1991;Marshall, 1989ab, 1990aM.E. Sereno, 1986M.E.…”