Objective validation of a dynamical and plausible computational model of visual attention

Abstract

Building a generic and highly capable vision system is still an open research problem. Actually, real-world vision systems need to face the challenge of dimensionality and ambiguity of data. To tackle this problem we introduced, in [1], a dynamic computational model of visual attention. This latter selects the most salient scene information while being able to adapt its behavior to the needs of a generic vision system. In this article, we focus on the objective validation of the plausibility of this original model. To check this property we compare (through three classical measures) the results obtained by several algorithms to an eye-tracking ground truth. Additionally, we study the influence of the model parameters on plausibility.