Multiscale saliency using natural statistics

Visual saliency is the perceptual quality that makes some items in visual scenes stand out from their immediate contexts. Natural scenes contain features across a large range of scales. However, most of the existing saliency models operate on a single scale. Inspired by the scale space theory and the saliency model using natural statistics, we propose a multiscale saliency model in this paper. It is demonstrated that the proposed model predicts well human gaze in free viewing static natural scenes.