{"id":115,"date":"2008-01-25T20:24:12","date_gmt":"2008-01-26T01:24:12","guid":{"rendered":"http:\/\/www.kayak2u.com\/blog\/?p=115"},"modified":"2008-01-25T20:24:12","modified_gmt":"2008-01-26T01:24:12","slug":"lotsa-little-pictures","status":"publish","type":"post","link":"http:\/\/www.kayak2u.com\/?p=115","title":{"rendered":"Lotsa little pictures"},"content":{"rendered":"<p>Did you ever want to see a <a target=\"_blank\" href=\"http:\/\/people.csail.mit.edu\/torralba\/tinyimages\/\">graphical representation of the semantic relationship of all the nouns in the English language<\/a>?&nbsp; &nbsp;Yup, there it is.&nbsp; There are 79 million or so images synthesized down and mapped to 53,463 nouns.&nbsp; It&#8217;s part of a project to train computers to recognize objects and scenes in images.&nbsp; The theory is, apparently, that if you have a hundred pictures of tables, and a hundred pictures of chairs, and you form a composite of the pictures by averaging them together, you&#8217;ve created a unique index of &quot;table&quot; and another unique index of &quot;chair&quot;.&nbsp; I have to suppose it will work in a general way, but if you click around the dataset there, you&#8217;ll see some interesting things (no pun intended).&nbsp; For entry #28543, which is the word &quot;hatemonger&quot;, it looks one of the pictures of the subset is that of Michael Moore, and another is of Lou Dobbs.&nbsp; Even without being told, one could guess that the image database was drawn from the internet.&nbsp; I was somewhat surprised when I clicked on the extreme right side of the next to bottom row, which turned out to be entry #53134 and the word &quot;bang&quot;.&nbsp; The banging that&#8217;s going on in those pictures is not what I would call a noun.<\/p>\n<p>via:&nbsp; <a target=\"_blank\" href=\"http:\/\/andrewsullivan.theatlantic.com\/the_daily_dish\/2008\/01\/eighty-million.html\">Andrew Sullivan<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Did you ever want to see a graphical representation of the semantic relationship of all the nouns in the English language?&nbsp; &nbsp;Yup, there it is.&nbsp; There are 79 million or so images synthesized down and mapped to 53,463 nouns.&nbsp; It&#8217;s part of a project to train computers to recognize objects and scenes in images.&nbsp; The [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"class_list":["post-115","post","type-post","status-publish","format-standard","hentry","category-misc"],"_links":{"self":[{"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=\/wp\/v2\/posts\/115","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=115"}],"version-history":[{"count":0,"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=\/wp\/v2\/posts\/115\/revisions"}],"wp:attachment":[{"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=115"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=115"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.kayak2u.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=115"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}