Network symmetry and binocular rivalry experiments.

Area of Science:

• Computational neuroscience
• Visual perception
• Cognitive modeling

Background:

• Hugh Wilson's models conceptualize decision-making as attribute-level network competition.
• Binocular rivalry presents surprising perceptual phenomena that require explanation.
• Existing models may not fully capture the complexities of visual competition.

Purpose of the Study:

• To propose a novel computational model, rivalry networks, for explaining binocular rivalry.
• To demonstrate how symmetry-breaking Hopf bifurcation in these networks can algorithmically account for observed percepts.
• To test the model's predictions against established psychophysical experiments.

Main Methods:

• Modification and extension of Wilson networks to create 'rivalry networks' with varied attributes and couplings.
• Application of a symmetry-breaking Hopf bifurcation algorithm within these rivalry networks.
• Analysis of psychophysical data from binocular rivalry experiments (Kovács et al., Shevell and Hong, Suzuki and Grabowecky).

Main Results:

• The proposed algorithm successfully explains surprising percepts observed in binocular rivalry experiments.
• The model predicts distinct outcomes for three-dot versus four-dot visual experiments.
• Rivalry networks offer a mechanistic explanation for perceptual competition.

Conclusions:

• Symmetry-breaking Hopf bifurcation in rivalry networks provides a viable algorithmic approach to understanding binocular rivalry.
• The model's ability to predict differences in dot experiments highlights its explanatory power.
• This framework advances computational models of visual consciousness and decision-making.