Saliency Detection via A Graph Based Diffusion Model

Zhouqin He, Bo Jiang, Yun Xiao, Chris H. Q. Ding, Bin Luo

Abstract: This paper proposes a graph based diffusion method for image saliency detection problem by adopting random walk with restart (RWR) model. Our method begins with computing background and foreground priors respectively for the input image. Based on these priors, we then adopt RWR method to obtain more reasonable and accurate background and foreground measurements by further considering the local structure of image. At last, we combine both background and foreground measurements together to obtain a more accurate saliency estimation. Experimental evaluations on four benchmark datasets demonstrate the benefits and effectiveness of the proposed method.