A neural network model for stroboscopic alternative motion

Abstract

A neural network which models multistable perception is presented. The network consists of sensor and inner neurons. The dynamics is established by a stochastic neuronal dynamics, a formal Hebb-type coupling dynamics and a resource mechanism that corresponds to saturation effects in perception. From this a system of coupled differential equations is derived and analyzed Single stimuli are bound to exactly one percept, even in ambiguous situations where multistability occurs. The network exhibits discontinuous as well as continuous phase transitions and models various empirical findings, including the percepts of succession, alternative motion and simultaneity, the percept of oscillation is explained by oscillating percepts of a continuous phase transition.

The dynamical foundations of motion pattern formation: Stability, selective adaptation, and perceptual continuity

“…In addition, Giese (1999) has shown that, in simplified form, the same principles can be the basis for quantitative simulations of the results of psychophysical experiments based on the motion quartet. Although all these phenomena could potentially have been comprehensively addressed by earlier dynamical models (e.g., Bartsch & van Hemmen, 1997;Carmesin & Arndt, 1996;Ditzinger & Haken, 1995;Francis & Grossberg, 1996;Grossberg & Mingolla,1985;Grossberg & Rudd, 1989Kawamoto & Anderson, 1985;Koechlin et al, 1999;Nowlan & Sejnowski, 1995;Williams et al, 1986), this is the first time that it has in fact been done. The results of the dynamical simulations in this article have been contrasted with Ullman's (1979) minimal mapping theory for explaining how the visual system solves the motion correspondence problem.…”

Linking dynamical perceptual decisions at different levels of description in motion pattern formation: Computational simulations

“…If frame duration were substantially longer than the duration of motion detector activation, future-shaping inhibitory interactions would likely become a less effective basis for hysteresis, and slow-time-scale future-shaping excitatory interactions would potentially become a more important contributor to pattern stability and hysteresis. There have been many earlier dynamical models in which inhibition among motion detectors with different directional selectivity has been the basis for pattern formation (Bartsch & van Hemmen, 1997;Carmesin & Arndt, 1996;Hock et al, 2003;Kawamoto & Anderson, 1985;Williams, Phillips, & Sekuler, 1986;Wilson & Kim, 1994), and there have also been dynamical multilevel models entailing motion processing in both areas V1 and MT (Chey, Grossberg, & Mingolla, 1997;Grossberg, Mingolla, & Viswanathan, 2001). 4 However, other than Hock et al's (2003) model, none of these have identified future-shaping interactions as the basis for the stability of perceived patterns.…”

Linking dynamical perceptual decisions at different levels of description in motion pattern formation: Psychophysics

“…In addition, computational simulations for random cinematograms (Williams, Phillips, & Sekuler, 1986), plaids (Wilson & Kim, 1994), and the motion quartet (see, e.g., Bartsch & van Hemmen, 1997;Carmesin & Arndt, 1996;Hock et al, 2003) indicate that motion pattern formation depends on recursive inhibitory interactions among detectors with different directional selectivity. MT therefore is a likely site for pattern-level dynamical decisions (see, e.g., Stoner & Albright, 1992).…”