Stochastic sensitivity of a bistable energy model for visual perception
Modern trends in physiology, psychology and cognitive neuroscience suggest that noise is an essential component of brain functionality and self-organization. With adequate noise the brain as a complex dynamical system can easily access different ordered states and improve signal detection for decision-making by preventing deadlocks. Using a stochastic sensitivity function approach, we analyze how sensitive equilibrium points are to Gaussian noise in a bistable energy model often used for qualitative description of visual perception. The probability distribution of noise-induced transitions between two coexisting percepts is calculated at different noise intensity and system stability. Stochastic squeezing of the hysteresis range and its transition from positive (bistable regime) to negative (intermittency regime) are demonstrated as the noise intensity increases. The hysteresis is more sensitive to noise in the system with higher stability.
KeywordsVisual perception Bistable model Noise Hysteresis
PACS Nos.05.45.-a 05.40.Ca 87.10.Ed 87.10.Mn 87.18.Tt
- L Arnold Random Dynamical Systems (Berlin: Springer) (1998)Google Scholar
- N Kogo, A Galli and J Wagemans Vis. Res. 51 2085 (2011)Google Scholar
- L A Necker The London and Edinburgh Philosophical Magazine and Journal of Science 1 (5) 329 (1832)Google Scholar
- E Rubin Visuell Wahrgenommene Figuren (Copenhagen: Glydenalske Boghande) (1921)Google Scholar