{"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/deep-transfer-learning-with-joint-adaptation","title":"Deep Transfer Learning with Joint Adaptation Networks","arxiv_id":"1605.06636","date":"2016-05-21","proceeding":"ICML 2017 8","authors":["Mingsheng Long","Han Zhu","Jian-Min Wang","Michael. I. Jordan"],"abstract":"Deep networks have been successfully applied to learn transferable features\nfor adapting models from a source domain to a different target domain. In this\npaper, we present joint adaptation networks (JAN), which learn a transfer\nnetwork by aligning the joint distributions of multiple domain-specific layers\nacross domains based on a joint maximum mean discrepancy (JMMD) criterion.\nAdversarial training strategy is adopted to maximize JMMD such that the\ndistributions of the source and target domains are made more distinguishable.\nLearning can be performed by stochastic gradient descent with the gradients\ncomputed by back-propagation in linear-time. Experiments testify that our model\nyields state of the art results on standard datasets.","url_abs":"http://arxiv.org/abs/1605.06636v2","url_pdf":"http://arxiv.org/pdf/1605.06636v2.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":"deep-transfer-learning-with-joint-adaptation","repo_url":"https://github.com/thuml/Transfer-Learning-Library","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"multi-source-unsupervised-domain-adaptation","task_name":"Multi-Source Unsupervised Domain Adaptation"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[{"slug":"imageclef-da","name":"ImageCLEF-DA","full_name":"ImageCLEF-DA"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/domain-adaptation-on-hmdbfull-to-ucf","task":"Domain Adaptation","dataset":"HMDBfull-to-UCF","model":"JAN","rank_in_archive_order":3,"of":5,"metrics":{"Accuracy":"79.69"},"uses_additional_data":false},{"leaderboard":"/sota/domain-adaptation-on-ucf-to-hmdbfull","task":"Domain Adaptation","dataset":"UCF-to-HMDBfull","model":"JAN","rank_in_archive_order":3,"of":5,"metrics":{"Accuracy":"74.72"},"uses_additional_data":false},{"leaderboard":"/sota/domain-adaptation-on-visda2017","task":"Domain Adaptation","dataset":"VisDA2017","model":"JAN","rank_in_archive_order":28,"of":28,"metrics":{"Accuracy":"58.3"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-domain-adaptation-on-office-home","task":"Unsupervised Domain Adaptation","dataset":"Office-Home","model":"JAN [cite:ICML17JAN]","rank_in_archive_order":13,"of":20,"metrics":{"Accuracy":"76.8"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1605.06636","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}