An astable multivibrator model of binocular rivalry.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Electronic Engineering

Background:

• Binocular rivalry is a phenomenon where dissimilar images presented to each eye result in alternating perception.
• Existing reciprocal inhibition models have faced challenges in replicating the stochastic nature of binocular rivalry.

Purpose of the Study:

• To explore the behavior of a neural network model for binocular rivalry.
• To establish an analogy between this neural model and an electronic astable multivibrator circuit.
• To demonstrate that reciprocal inhibition models can reproduce experimentally observed properties of rivalry.

Main Methods:

• Development of a neural network model with reciprocal feedback inhibition between left and right eye signals.
• Analogy drawn between the neural model's behavior and an electronic astable multivibrator circuit.
• Computer simulations to test the model's ability to reproduce rivalry dynamics.

Main Results:

• The model's behavior (rivalrous oscillations or stable fusion) is determined by the strength of inhibitory coupling.
• Strong coupling leads to oscillations, while weak coupling results in stable fusion.
• Computer simulations confirm the model's capacity to reproduce the stochastic behavior of binocular rivalry.

Conclusions:

• The proposed neural network model, analogous to an astable multivibrator, successfully simulates binocular rivalry.
• Inhibitory coupling strength is a critical factor influencing the transition between rivalry and fusion.
• This model serves as a counterexample to claims limiting the capabilities of reciprocal inhibition models in explaining rivalry.