{"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-object-centric-representations-for","title":"Deep Object-Centric Representations for Generalizable Robot Learning","arxiv_id":"1708.04225","date":"2017-08-14","proceeding":null,"authors":["Coline Devin","Pieter Abbeel","Trevor Darrell","Sergey Levine"],"abstract":"Robotic manipulation in complex open-world scenarios requires both reliable\nphysical manipulation skills and effective and generalizable perception. In\nthis paper, we propose a method where general purpose pretrained visual models\nserve as an object-centric prior for the perception system of a learned policy.\nWe devise an object-level attentional mechanism that can be used to determine\nrelevant objects from a few trajectories or demonstrations, and then\nimmediately incorporate those objects into a learned policy. A task-independent\nmeta-attention locates possible objects in the scene, and a task-specific\nattention identifies which objects are predictive of the trajectories. The\nscope of the task-specific attention is easily adjusted by showing\ndemonstrations with distractor objects or with diverse relevant objects. Our\nresults indicate that this approach exhibits good generalization across object\ninstances using very few samples, and can be used to learn a variety of\nmanipulation tasks using reinforcement learning.","url_abs":"http://arxiv.org/abs/1708.04225v3","url_pdf":"http://arxiv.org/pdf/1708.04225v3.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-object-centric-representations-for","repo_url":"https://github.com/cdevin/objectattention","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"object","task_name":"Object"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.04225","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}