{"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/end-to-end-representation-learning-for","title":"End-to-end representation learning for Correlation Filter based tracking","arxiv_id":"1704.06036","date":"2017-04-20","proceeding":"CVPR 2017 7","authors":["Jack Valmadre","Luca Bertinetto","João F. Henriques","Andrea Vedaldi","Philip H. S. Torr"],"abstract":"The Correlation Filter is an algorithm that trains a linear template to\ndiscriminate between images and their translations. It is well suited to object\ntracking because its formulation in the Fourier domain provides a fast\nsolution, enabling the detector to be re-trained once per frame. Previous works\nthat use the Correlation Filter, however, have adopted features that were\neither manually designed or trained for a different task. This work is the\nfirst to overcome this limitation by interpreting the Correlation Filter\nlearner, which has a closed-form solution, as a differentiable layer in a deep\nneural network. This enables learning deep features that are tightly coupled to\nthe Correlation Filter. Experiments illustrate that our method has the\nimportant practical benefit of allowing lightweight architectures to achieve\nstate-of-the-art performance at high framerates.","url_abs":"http://arxiv.org/abs/1704.06036v1","url_pdf":"http://arxiv.org/pdf/1704.06036v1.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":"object-tracking","task_name":"Object Tracking"},{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-object-tracking-on-otb-2013","task":"Visual Object Tracking","dataset":"OTB-2013","model":"CFNet","rank_in_archive_order":6,"of":7,"metrics":{"AUC":"0.611"},"uses_additional_data":false},{"leaderboard":"/sota/visual-object-tracking-on-otb-2015","task":"Visual Object Tracking","dataset":"OTB-2015","model":"CFNet","rank_in_archive_order":17,"of":18,"metrics":{"AUC":"0.568"},"uses_additional_data":false},{"leaderboard":"/sota/visual-object-tracking-on-otb-50","task":"Visual Object Tracking","dataset":"OTB-50","model":"CFNet","rank_in_archive_order":3,"of":4,"metrics":{"AUC":"0.530"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1704.06036","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}