{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/instance-level-salient-object-segmentation","title":"Instance-Level Salient Object Segmentation","arxiv_id":"1704.03604","date":"2017-04-12","proceeding":"CVPR 2017 7","authors":["Guanbin Li","Yuan Xie","Liang Lin","Yizhou Yu"],"abstract":"Image saliency detection has recently witnessed rapid progress due to deep\nconvolutional neural networks. However, none of the existing methods is able to\nidentify object instances in the detected salient regions. In this paper, we\npresent a salient instance segmentation method that produces a saliency mask\nwith distinct object instance labels for an input image. Our method consists of\nthree steps, estimating saliency map, detecting salient object contours and\nidentifying salient object instances. For the first two steps, we propose a\nmultiscale saliency refinement network, which generates high-quality salient\nregion masks and salient object contours. Once integrated with multiscale\ncombinatorial grouping and a MAP-based subset optimization framework, our\nmethod can generate very promising salient object instance segmentation\nresults. To promote further research and evaluation of salient instance\nsegmentation, we also construct a new database of 1000 images and their\npixelwise salient instance annotations. Experimental results demonstrate that\nour proposed method is capable of achieving state-of-the-art performance on all\npublic benchmarks for salient region detection as well as on our new dataset\nfor salient instance segmentation.","url_abs":"http://arxiv.org/abs/1704.03604v1","url_pdf":"http://arxiv.org/pdf/1704.03604v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"instance-segmentation","task_name":"Instance Segmentation"},{"task_slug":"object","task_name":"Object"},{"task_slug":"saliency-detection","task_name":"Saliency Detection"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/salient-object-detection-on-duts-te","task":"RGB Salient Object Detection","dataset":"DUTS-TE","model":"MSR","rank_in_archive_order":22,"of":31,"metrics":{"MAE":"0.062","max F-measure":"0.824"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1704.03604","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}