{"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/amulet-aggregating-multi-level-convolutional","title":"Amulet: Aggregating Multi-level Convolutional Features for Salient Object Detection","arxiv_id":"1708.02001","date":"2017-08-07","proceeding":"ICCV 2017 10","authors":["Pingping Zhang","Dong Wang","Huchuan Lu","Hongyu Wang","Xiang Ruan"],"abstract":"Fully convolutional neural networks (FCNs) have shown outstanding performance\nin many dense labeling problems. One key pillar of these successes is mining\nrelevant information from features in convolutional layers. However, how to\nbetter aggregate multi-level convolutional feature maps for salient object\ndetection is underexplored. In this work, we present Amulet, a generic\naggregating multi-level convolutional feature framework for salient object\ndetection. Our framework first integrates multi-level feature maps into\nmultiple resolutions, which simultaneously incorporate coarse semantics and\nfine details. Then it adaptively learns to combine these feature maps at each\nresolution and predict saliency maps with the combined features. Finally, the\npredicted results are efficiently fused to generate the final saliency map. In\naddition, to achieve accurate boundary inference and semantic enhancement,\nedge-aware feature maps in low-level layers and the predicted results of low\nresolution features are recursively embedded into the learning framework. By\naggregating multi-level convolutional features in this efficient and flexible\nmanner, the proposed saliency model provides accurate salient object labeling.\nComprehensive experiments demonstrate that our method performs favorably\nagainst state-of-the art approaches in terms of near all compared evaluation\nmetrics.","url_abs":"http://arxiv.org/abs/1708.02001v1","url_pdf":"http://arxiv.org/pdf/1708.02001v1.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":[{"paper_slug":"amulet-aggregating-multi-level-convolutional","repo_url":"https://github.com/Pchank/caffe-sal","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"salient-object-detection","task_name":"RGB Salient Object Detection"},{"task_slug":"salient-object-detection-1","task_name":"Salient Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/salient-object-detection-on-duts-te","task":"RGB Salient Object Detection","dataset":"DUTS-TE","model":"Amulet","rank_in_archive_order":27,"of":31,"metrics":{"MAE":"0.075","max F-measure":"0.773"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.02001","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}