{"ID":13790,"CreatedAt":"2026-02-27T13:00:40Z","UpdatedAt":"2026-02-27T13:00:40Z","DeletedAt":null,"paper_url":"https://paperswithcode.com/paper/a-novel-graph-structure-for-salient-object","arxiv_id":"1711.11266","title":"A novel graph structure for salient object detection based on divergence background and compact foreground","abstract":"In this paper, we propose an efficient and discriminative model for salient\nobject detection. Our method is carried out in a stepwise mechanism based on\nboth divergence background and compact foreground cues. In order to effectively\nenhance the distinction between nodes along object boundaries and the\nsimilarity among object regions, a graph is constructed by introducing the\nconcept of virtual node. To remove incorrect outputs, a scheme for selecting\nbackground seeds and a method for generating compactness foreground regions are\nintroduced, respectively. Different from prior methods, we calculate the\nsaliency value of each node based on the relationship between the corresponding\nnode and the virtual node. In order to achieve significant performance\nimprovement consistently, we propose an Extended Manifold Ranking (EMR)\nalgorithm, which subtly combines suppressed / active nodes and mid-level\ninformation. Extensive experimental results demonstrate that the proposed\nalgorithm performs favorably against the state-of-art saliency detection\nmethods in terms of different evaluation metrics on several benchmark datasets.","url_abs":"http://arxiv.org/abs/1711.11266v1","url_pdf":"http://arxiv.org/pdf/1711.11266v1.pdf","authors":"[\"Chenxing Xia\", \"Hanling Zhang\", \"Keqin Li\"]","published":"2017-11-30T00:00:00Z","tasks":"[\"Object\", \"object-detection\", \"Object Detection\", \"RGB Salient Object Detection\", \"Saliency Detection\", \"Salient Object Detection\"]","methods":"[]","has_code":false}