{"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/student-becoming-the-master-knowledge","title":"Student Becoming the Master: Knowledge Amalgamation for Joint Scene Parsing, Depth Estimation, and More","arxiv_id":"1904.10167","date":"2019-04-23","proceeding":"CVPR 2019 6","authors":["Jingwen Ye","Yixin Ji","Xinchao Wang","Kairi Ou","Dapeng Tao","Mingli Song"],"abstract":"In this paper, we investigate a novel deep-model reusing task. Our goal is to\ntrain a lightweight and versatile student model, without human-labelled\nannotations, that amalgamates the knowledge and masters the expertise of two\npretrained teacher models working on heterogeneous problems, one on scene\nparsing and the other on depth estimation. To this end, we propose an\ninnovative training strategy that learns the parameters of the student\nintertwined with the teachers, achieved by 'projecting' its amalgamated\nfeatures onto each teacher's domain and computing the loss. We also introduce\ntwo options to generalize the proposed training strategy to handle three or\nmore tasks simultaneously. The proposed scheme yields very encouraging results.\nAs demonstrated on several benchmarks, the trained student model achieves\nresults even superior to those of the teachers in their own expertise domains\nand on par with the state-of-the-art fully supervised models relying on\nhuman-labelled annotations.","url_abs":"http://arxiv.org/abs/1904.10167v1","url_pdf":"http://arxiv.org/pdf/1904.10167v1.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":"student-becoming-the-master-knowledge","repo_url":"https://github.com/zju-vipa/KamalEngine","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"scene-parsing","task_name":"Scene Parsing"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.10167","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}