{"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/exploiting-web-images-for-weakly-supervised","title":"Exploiting Web Images for Weakly Supervised Object Detection","arxiv_id":"1707.08721","date":"2017-07-27","proceeding":null,"authors":["Qingyi Tao","Hao Yang","Jianfei Cai"],"abstract":"In recent years, the performance of object detection has advanced\nsignificantly with the evolving deep convolutional neural networks. However,\nthe state-of-the-art object detection methods still rely on accurate bounding\nbox annotations that require extensive human labelling. Object detection\nwithout bounding box annotations, i.e, weakly supervised detection methods, are\nstill lagging far behind. As weakly supervised detection only uses image level\nlabels and does not require the ground truth of bounding box location and label\nof each object in an image, it is generally very difficult to distill knowledge\nof the actual appearances of objects. Inspired by curriculum learning, this\npaper proposes an easy-to-hard knowledge transfer scheme that incorporates easy\nweb images to provide prior knowledge of object appearance as a good starting\npoint. While exploiting large-scale free web imagery, we introduce a\nsophisticated labour free method to construct a web dataset with good diversity\nin object appearance. After that, semantic relevance and distribution relevance\nare introduced and utilized in the proposed curriculum training scheme. Our\nend-to-end learning with the constructed web data achieves remarkable\nimprovement across most object classes especially for the classes that are\noften considered hard in other works.","url_abs":"http://arxiv.org/abs/1707.08721v2","url_pdf":"http://arxiv.org/pdf/1707.08721v2.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":"diversity","task_name":"Diversity"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"},{"task_slug":"weakly-supervised-object-detection","task_name":"Weakly Supervised Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/weakly-supervised-object-detection-on-pascal-1","task":"Weakly Supervised Object Detection","dataset":"PASCAL VOC 2007","model":"WebRelETH","rank_in_archive_order":30,"of":41,"metrics":{"MAP":"46.0"},"uses_additional_data":true},{"leaderboard":"/sota/weakly-supervised-object-detection-on-pascal","task":"Weakly Supervised Object Detection","dataset":"PASCAL VOC 2012 test","model":"WebRelETH","rank_in_archive_order":23,"of":32,"metrics":{"MAP":"42.8"},"uses_additional_data":true}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}