Oscillations in the perception of ambiguous patterns a model based on synergetics

Abstract

Abstract. An algorithm which can be realized by synergetic systems and which was introduced by one of us (Haken 1987) for the recognition of patterns is extended so that the perception of ambiguous patterns can be modelled. In this approach so-called attention parameters are subjected to a damping mechanism mimicking the effect of saturation of attention. In this way oscillations of perception arise quite naturally. Our approach takes also ambiguous patterns with bias into account which leads to different periods of the attention paid to the one or the other interpretation of the pattern. Our results are in good agreement with previous psycho-physical studies by other authors. Finally we show how hysteresis of perception can be modelled.

A nonlinear dynamics model for simulating long range correlations of cognitive bistability

“…To the best of our knowledge, Ditzinger and Haken (1989) were the first who transferred the dynamics to the macroscopic behavioral level of perception state, attention, and memory control parameters using the so-called slaving principle of synergetics (theory of self-organizing dynamical systems): macroscopic-order parameters are created through the nonlinear interaction of the neural microsystems and, in turn, slave the microscopic behavior into coherent synchronous activity (Haken 1978). Within the framework of synergetics, the dynamics of complex systems are often governed by a few macroscopic-order parameters only.…”

“…Like in Ditzinger and Haken (1989), the basic model couples the dynamics of a macroscopic (behavioral) perception-state-order parameter with an adaptive attention (feedback gain) control parameter with additive noise (Fürstenau 2003(Fürstenau , 2004. The order parameter dynamics is derived from a radically simplified field equation with phase feedback of superimposed fields which formalize the potential percepts.…”

“…Concerning the interpretation of experimental results of bistable perception, there is an ongoing controversial discussion on the possible deterministic, even chaotic or fractal dynamics of cognitive bistability (Poston and Stewart 1978;Ditzinger and Haken 1989;Wilson 1999;Richards et al 1994;Fürstenau 2004;Noest et al 2007) versus dominance of stochastic properties (De Marco et al 1977;Lehky 1995;Merk and Schnakenberg 2002), as derived from statistical analysis of reversal time series. From the above given overview, it appears reasonable to assume that both deterministic and stochastic contributions to the cognitive system dynamics have to be considered.…”

“…As a kind of minimum architecture allowing for the emergence of fast quasi-discontinuous state transitions, reentrant perception-attention dynamics with attention fatigue (Orbach et al 1963;Ditzinger and Haken 1989) and delayed feedback is employed (Fürstenau 2004(Fürstenau , 2006, with delay T and feedback gain g. In the present model, a macroscopic behavioral approach is used with a dynamical perception-state-order parameter v, an adaptive attention control parameter G ∼ g, and a dynamical perception bias parameter v b representing short term memory and learning. I will motivate this architecture by relating the three dynamical model parameters to thalamocortical reentrant loops as proposed by Tononi and Edelman (1998) and to the bottom-up and top-down aspects of visual attention (Itti and Koch 2001).…”

“…Consequently, in the present model, the feedback gain serves as adaptive control parameter which induces the rapid transitions between the alternative stationary perception states P1 and P2 through attention fatigue (Orbach et al 1963), which is modeled according to Ditzinger and Haken (1989) via coupling to the perception state v.…”