Saliency optimization via low rank matrix recovery with multi-prior integration

Abstract

In this paper, we propose an unsupervised saliency detection method based on the low rank matrix recovery model (LRMR). Learning of a feature transform matrix is not needed and three low level priors are integrated in our model in replacing of the original high level ones, which could act as better guidance cues. Also an optimization framework is designed to optimize the raw saliency map generated by the low rank matrix recovery model. We compare our method with seven previous methods and test them on several benchmark datasets. The results demonstrate that our model achieves state of the art performance.