{"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/leveraging-unlabeled-data-for-crowd-counting","title":"Leveraging Unlabeled Data for Crowd Counting by Learning to Rank","arxiv_id":"1803.03095","date":"2018-03-08","proceeding":"CVPR 2018 6","authors":["Xialei Liu","Joost Van de Weijer","Andrew D. Bagdanov"],"abstract":"We propose a novel crowd counting approach that leverages abundantly\navailable unlabeled crowd imagery in a learning-to-rank framework. To induce a\nranking of cropped images , we use the observation that any sub-image of a\ncrowded scene image is guaranteed to contain the same number or fewer persons\nthan the super-image. This allows us to address the problem of limited size of\nexisting datasets for crowd counting. We collect two crowd scene datasets from\nGoogle using keyword searches and query-by-example image retrieval,\nrespectively. We demonstrate how to efficiently learn from these unlabeled\ndatasets by incorporating learning-to-rank in a multi-task network which\nsimultaneously ranks images and estimates crowd density maps. Experiments on\ntwo of the most challenging crowd counting datasets show that our approach\nobtains state-of-the-art results.","url_abs":"http://arxiv.org/abs/1803.03095v1","url_pdf":"http://arxiv.org/pdf/1803.03095v1.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":"leveraging-unlabeled-data-for-crowd-counting","repo_url":"https://github.com/xialeiliu/CrowdCountingCVPR18","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"crowd-counting","task_name":"Crowd Counting"},{"task_slug":"image-retrieval","task_name":"Image Retrieval"},{"task_slug":"learning-to-rank","task_name":"Learning-To-Rank"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/crowd-counting-on-shanghaitech-a","task":"Crowd Counting","dataset":"ShanghaiTech A","model":"Liu et al.","rank_in_archive_order":29,"of":35,"metrics":{"MAE":"73.6"},"uses_additional_data":false},{"leaderboard":"/sota/crowd-counting-on-shanghaitech-b","task":"Crowd Counting","dataset":"ShanghaiTech B","model":"Liu et al.","rank_in_archive_order":25,"of":32,"metrics":{"MAE":"13.7"},"uses_additional_data":false},{"leaderboard":"/sota/crowd-counting-on-ucf-cc-50","task":"Crowd Counting","dataset":"UCF CC 50","model":"Liu et al.","rank_in_archive_order":19,"of":22,"metrics":{"MAE":"337.6"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.03095","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}