Competition for visual representation is an important mechanism for selective visual attention. The traditionalglobal distinctiveness based saliency models usually compute the distinctiveness to measure saliency via comparingthe di erence of image patches in various spaces. In this paper, we propose to use an improved neuralcompetition model to replace the comparison. The pairwise competition responses for a patch to all of the otherpatches are summed up to represent the distinctiveness of that patch. Particularly, the competition responseis computed by a neural competition model with the dissimilarity bias and the gradient based feature inputs.Experimental results validate that the proposed model presents high e ectiveness in saliency detection byoutperforming nine state-of-the-art models.
展开▼
