{"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-mutual-learning","title":"Deep Mutual Learning","arxiv_id":"1706.00384","date":"2017-06-01","proceeding":"CVPR 2018 6","authors":["Ying Zhang","Tao Xiang","Timothy M. Hospedales","Huchuan Lu"],"abstract":"Model distillation is an effective and widely used technique to transfer\nknowledge from a teacher to a student network. The typical application is to\ntransfer from a powerful large network or ensemble to a small network, that is\nbetter suited to low-memory or fast execution requirements. In this paper, we\npresent a deep mutual learning (DML) strategy where, rather than one way\ntransfer between a static pre-defined teacher and a student, an ensemble of\nstudents learn collaboratively and teach each other throughout the training\nprocess. Our experiments show that a variety of network architectures benefit\nfrom mutual learning and achieve compelling results on CIFAR-100 recognition\nand Market-1501 person re-identification benchmarks. Surprisingly, it is\nrevealed that no prior powerful teacher network is necessary -- mutual learning\nof a collection of simple student networks works, and moreover outperforms\ndistillation from a more powerful yet static teacher.","url_abs":"http://arxiv.org/abs/1706.00384v1","url_pdf":"http://arxiv.org/pdf/1706.00384v1.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-mutual-learning","repo_url":"https://github.com/abhishirk/Aligned_ReId","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/aquvitae/aquvitae","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/generation21/generation6011","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/h4veFunCodin9/Aligned_ReID","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/huanghoujing/AlignedReID-Re-Production-Pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/pilsHan/DML","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/pilsHan/DML_for-personal-study","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"deep-mutual-learning","repo_url":"https://github.com/shubhamtyagii/Aligned_Reid","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"person-re-identification","task_name":"Person Re-Identification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1706.00384","atlas_url":"https://app.syntology.ai/?focus=1706.00384","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}