{"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/visual-reinforcement-learning-with-imagined","title":"Visual Reinforcement Learning with Imagined Goals","arxiv_id":"1807.04742","date":"2018-07-12","proceeding":"NeurIPS 2018 12","authors":["Ashvin Nair","Vitchyr Pong","Murtaza Dalal","Shikhar Bahl","Steven Lin","Sergey Levine"],"abstract":"For an autonomous agent to fulfill a wide range of user-specified goals at\ntest time, it must be able to learn broadly applicable and general-purpose\nskill repertoires. Furthermore, to provide the requisite level of generality,\nthese skills must handle raw sensory input such as images. In this paper, we\npropose an algorithm that acquires such general-purpose skills by combining\nunsupervised representation learning and reinforcement learning of\ngoal-conditioned policies. Since the particular goals that might be required at\ntest-time are not known in advance, the agent performs a self-supervised\n\"practice\" phase where it imagines goals and attempts to achieve them. We learn\na visual representation with three distinct purposes: sampling goals for\nself-supervised practice, providing a structured transformation of raw sensory\ninputs, and computing a reward signal for goal reaching. We also propose a\nretroactive goal relabeling scheme to further improve the sample-efficiency of\nour method. Our off-policy algorithm is efficient enough to learn policies that\noperate on raw image observations and goals for a real-world robotic system,\nand substantially outperforms prior techniques.","url_abs":"http://arxiv.org/abs/1807.04742v2","url_pdf":"http://arxiv.org/pdf/1807.04742v2.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":"visual-reinforcement-learning-with-imagined","repo_url":"https://github.com/vitchyr/rlkit","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"visual-reinforcement-learning-with-imagined","repo_url":"https://github.com/vitchyr/multiworld","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.04742","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}