An object-based visual selection model combining physical features and memory

Abstract

In this paper, a new visual selection model is proposed, which combines both early visual features and object-based visual selection modulations. This model integrates three main mechanisms. The first is responsible for the segmentation of the scene allowing the identification of objects. In the second one, the average of saliency of each object is calculated for each feature considered in this work, which provides the modulation of the visual attention for one or more features. Finally, the third mechanism is responsible for building the object-saliency map, which highlights the salient objects in the scene. It will be shown that top-down modulation can overcome bottom-up saliency by selecting a known object instead of the most salient (bottom-up) and is even clear in the absence of any bottom-up clue. Several experiments with synthetic and real images are conducted and the obtained results demonstrate the effectiveness of the proposed approach for visual attention.