{"id":57405,"date":"2017-05-16T11:23:30","date_gmt":"2017-05-16T18:23:30","guid":{"rendered":"https:\/\/lifeboat.com\/blog\/2017\/05\/robots-that-learn"},"modified":"2017-05-23T15:01:39","modified_gmt":"2017-05-23T22:01:39","slug":"robots-that-learn","status":"publish","type":"post","link":"https:\/\/lifeboat.com\/blog\/2017\/05\/robots-that-learn","title":{"rendered":"Robots that Learn"},"content":{"rendered":"<p><a class=\"aligncenter blog-photo\" href=\"https:\/\/lifeboat.com\/blog.images\/robots-that-learn.gif\"><\/a><\/p>\n<p>Last month, we showed an <a href=\"https:\/\/blog.openai.com\/spam-detection-in-the-physical-world\/\">earlier version<\/a> of this robot where we\u2019d trained its vision system using <a href=\"https:\/\/arxiv.org\/abs\/1703.06907\">domain randomization<\/a>, that is, by showing it simulated objects with a variety of color, backgrounds, and textures, without the use of any real images.<\/p>\n<p>Now, we\u2019ve developed and deployed a new algorithm, <a href=\"https:\/\/arxiv.org\/abs\/1703.07326\">one-shot imitation learning<\/a>, allowing a human to communicate how to do a new task by performing it in <a href=\"https:\/\/en.wikipedia.org\/wiki\/Virtual_reality\">VR<\/a>. Given a single demonstration, the robot is able to solve the same task from an arbitrary starting configuration.<\/p>\n<p><img decoding=\"async\" style=\"display:none;\" src=\"https:\/\/lifeboat.com\/blog.images\/robots-that-learn.jpg\" alt=\"\" \/> <em>Caption: Our system can learn a behavior from a single demonstration delivered within a simulator, then reproduce that behavior in different setups in reality.<\/em><\/p>\n<p><!-- Link: <a href=\"https:\/\/blog.openai.com\/robots-that-learn\/\">https:\/\/blog.openai.com\/robots-that-learn\/<\/a> --><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Last month, we showed an earlier version of this robot where we\u2019d trained its vision system using domain randomization, that is, by showing it simulated objects with a variety of color, backgrounds, and textures, without the use of any real images. Now, we\u2019ve developed and deployed a new algorithm, one-shot imitation learning, allowing a human [\u2026]<\/p>\n","protected":false},"author":354,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[41,6,1879],"tags":[],"class_list":["post-57405","post","type-post","status-publish","format-standard","hentry","category-information-science","category-robotics-ai","category-virtual-reality"],"_links":{"self":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/57405","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/users\/354"}],"replies":[{"embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/comments?post=57405"}],"version-history":[{"count":1,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/57405\/revisions"}],"predecessor-version":[{"id":57532,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/posts\/57405\/revisions\/57532"}],"wp:attachment":[{"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/media?parent=57405"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/categories?post=57405"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/lifeboat.com\/blog\/wp-json\/wp\/v2\/tags?post=57405"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}