A neural network model of the attentional blink phenomenon

Abstract

The Attentional Blink refers to the finding that the perception of a target stimulus is often impaired when it is presented within a temporal window of 200-500 ms following another successfully-perceived stimulus. This interesting phenomenon has attracted much research in the fields of attention and perception but has also inspired the implementation of several computational models. In the present study we have implemented a model of selective attention that is capable of modelling a variety of findings related to the Attentional Blink. Importantly, the model produces synchronization of neural activity to simulate the interaction between a low-level visual system and a high-level goal-maintenance system during the deployment of attention. Synchronization of neural activity within and across brain areas is typically observed when carrying out various tasks that involve the attentional processing of information. Thus, the model provides a neurally and computationally plausible account for the Attentional Blink and potentially other attentional tasks.