{"id":5020,"date":"2017-02-04T07:36:01","date_gmt":"2017-02-04T15:36:01","guid":{"rendered":"http:\/\/jnack.com\/blog\/?p=5020"},"modified":"2017-02-03T16:36:56","modified_gmt":"2017-02-04T00:36:56","slug":"enhance-imaging-tech-makes-something-out-of-almost-nothing","status":"publish","type":"post","link":"http:\/\/jnack.com\/blog\/2017\/02\/04\/enhance-imaging-tech-makes-something-out-of-almost-nothing\/","title":{"rendered":"&#8220;Enhance!&#8221;: Imaging tech makes something out of (almost) nothing"},"content":{"rendered":"<p>Wow\u2014<a href=\"https:\/\/arxiv.org\/abs\/1702.00783\">this paper<\/a>\u00a0(don\u2019t worry, I\u2019m not going to read it either) promises to recreate face data from <em>extremely<\/em> low-res images. As Yonatan Zunger <a href=\"https:\/\/plus.google.com\/+YonatanZunger\/posts\/fEKccXfJcgr?deb=oz%3A3%3Aall&amp;opt=pssrw%3Aexp.mevin_use_global_profile_for_fallback%3Atrue%2Cpssrw%3Aexp.enable_mevin%3Atrue%2Cpssrw%3Aexp.enable_sherlock%3Afalse\">explains<\/a>,<\/p>\n<blockquote>\n<p>[I]t takes a pixelated image, and uses the fact that it knows it&#8217;s looking at a human face, and what human faces look like, to turn each pixel into a 4&#215;4 grid of its best guess of which colors would have to have been there to both be consistent with a face shape and with the average color it saw.<\/p>\n<p>On the right are the original pictures, at 32&#215;32 resolution. On the left is what happens after they&#8217;re reduced down to 8&#215;8, the sort of thing you would get when a camera is at the limit of its resolution. In the middle is what their algorithm recovered.<\/p>\n<\/blockquote>\n<p><img decoding=\"async\" loading=\"lazy\" title=\"NewImage.png\" src=\"http:\/\/jnack.com\/blog\/wp-content\/uploads\/2017\/02\/NewImage-3.png\" alt=\"NewImage\" width=\"535\" height=\"600\" border=\"0\" \/><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Wow\u2014this paper\u00a0(don\u2019t worry, I\u2019m not going to read it either) promises to recreate face data from extremely low-res images. As Yonatan Zunger explains, [I]t takes a pixelated image, and uses the fact that it knows it&#8217;s looking at a human face, and what human faces look like, to turn each pixel into a 4&#215;4 grid [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[3],"tags":[],"_links":{"self":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/5020"}],"collection":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/comments?post=5020"}],"version-history":[{"count":3,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/5020\/revisions"}],"predecessor-version":[{"id":5023,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/5020\/revisions\/5023"}],"wp:attachment":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/media?parent=5020"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/categories?post=5020"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/tags?post=5020"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}