An agent cognitive model for visual attention and response to novelty

Abstract

Cognitive virtual agents are useful in human behavior simulation. We present a biologically inspired cognitive model for visual attention that takes into account the occurrence of novel stimulus, and it deals with the habituation to novelty. Our approach relies on the identification of cerebral areas involved in attention, semantic memory and non-associative learning; the processes related to each of them and the hypothetical information generated in each step. The model described in this paper is capable to be integrated in a cognitive architecture to interact with other cognitive functions.