A Learned Saliency Predictor for Dynamic Natural Scenes

Abstract. We investigate the extent to which eye movements in natural dynamic scenes can be predicted with a simple model of bottom-up saliency, which learns on different visual representations to discriminate between salient and less salient movie regions. Our image representations, the geometrical invariants of the structure tensor, are computed on multiple scales of an anisotropic spatio-temporal multiresolution pyramid. Eye movement data is used to label video locations as salient. For each location, low-dimensional features are extracted on the multiscale representations and used to train a classifier. The quality of the predictor is tested on a large test set of eye movement data and compared with the performance of two state-of-the-art saliency models on this data set. The proposed model demonstrates significant improvement

• Type: Conference
• Year: 2010
• Where: ICANN
• Authors: Eleonora Vig, Michael Dorr, Thomas Martinetz, Erhardt Barth
• Keywords: Eye Movement, Eye Movement Data, ICANN 2010, Neural Networks, Salient Movie Regions