{"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/supervision-via-competition-robot-adversaries","title":"Supervision via Competition: Robot Adversaries for Learning Tasks","arxiv_id":"1610.01685","date":"2016-10-05","proceeding":null,"authors":["Lerrel Pinto","James Davidson","Abhinav Gupta"],"abstract":"There has been a recent paradigm shift in robotics to data-driven learning\nfor planning and control. Due to large number of experiences required for\ntraining, most of these approaches use a self-supervised paradigm: using\nsensors to measure success/failure. However, in most cases, these sensors\nprovide weak supervision at best. In this work, we propose an adversarial\nlearning framework that pits an adversary against the robot learning the task.\nIn an effort to defeat the adversary, the original robot learns to perform the\ntask with more robustness leading to overall improved performance. We show that\nthis adversarial framework forces the the robot to learn a better grasping\nmodel in order to overcome the adversary. By grasping 82% of presented novel\nobjects compared to 68% without an adversary, we demonstrate the utility of\ncreating adversaries. We also demonstrate via experiments that having robots in\nadversarial setting might be a better learning strategy as compared to having\ncollaborative multiple robots.","url_abs":"http://arxiv.org/abs/1610.01685v1","url_pdf":"http://arxiv.org/pdf/1610.01685v1.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":"supervision-via-competition-robot-adversaries","repo_url":"https://github.com/hudongrui/2018SU_Frank_Zihan","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1610.01685","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}