{"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/dari-distance-metric-and-representation","title":"DARI: Distance metric And Representation Integration for Person Verification","arxiv_id":"1604.04377","date":"2016-04-15","proceeding":null,"authors":["Guangrun Wang","Liang Lin","Shengyong Ding","Ya Li","Qing Wang"],"abstract":"The past decade has witnessed the rapid development of feature representation\nlearning and distance metric learning, whereas the two steps are often\ndiscussed separately. To explore their interaction, this work proposes an\nend-to-end learning framework called DARI, i.e. Distance metric And\nRepresentation Integration, and validates the effectiveness of DARI in the\nchallenging task of person verification. Given the training images annotated\nwith the labels, we first produce a large number of triplet units, and each one\ncontains three images, i.e. one person and the matched/mismatch references. For\neach triplet unit, the distance disparity between the matched pair and the\nmismatched pair tends to be maximized. We solve this objective by building a\ndeep architecture of convolutional neural networks. In particular, the\nMahalanobis distance matrix is naturally factorized as one top fully-connected\nlayer that is seamlessly integrated with other bottom layers representing the\nimage feature. The image feature and the distance metric can be thus\nsimultaneously optimized via the one-shot backward propagation. On several\npublic datasets, DARI shows very promising performance on re-identifying\nindividuals cross cameras against various challenges, and outperforms other\nstate-of-the-art approaches.","url_abs":"http://arxiv.org/abs/1604.04377v1","url_pdf":"http://arxiv.org/pdf/1604.04377v1.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":"metric-learning","task_name":"Metric Learning"},{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":null,"task_name":"Triplet"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/person-re-identification-on-sysu-30k","task":"Person Re-Identification","dataset":"SYSU-30k","model":"DARI (generalization)","rank_in_archive_order":7,"of":10,"metrics":{" Rank-1":"11.2"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/1604.04377","atlas_url":"https://app.syntology.ai/?focus=1604.04377","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}