Object of interest detection by saliency learning

摘要: In this paper, we present a method for object of interest detection. This is statistical in nature and hinges model which combines salient features using mixture linear support vector machines. It exploits divide-and-conquer strategy by partitioning the feature space into sub-regions linearly separable data-points. yields structured learning approach where learn machine each region, weights, combination parameters at hand. Thus, learns such that classifiers can be used to recover objects image. We illustrate utility applying our algorithm MSRA Salient Object Database.