{"id":1105,"date":"2013-02-05T13:35:39","date_gmt":"2013-02-05T11:35:39","guid":{"rendered":"http:\/\/wordpress.khm.de\/sag\/?p=1105"},"modified":"2013-02-05T13:35:39","modified_gmt":"2013-02-05T11:35:39","slug":"anti-facial-recognition-strategies-defeated","status":"publish","type":"post","link":"https:\/\/sag.khm.de\/?p=1105","title":{"rendered":"Anti-Facial Recognition strategies defeated"},"content":{"rendered":"<p>We had some fun at my &#8216;Raumkontrolle&#8217; block seminar, trying out our Netavis Observer &#8216;face recognition&#8217; software on various anti-Anti-Facial Recognition strategies. Seems they all used the same (Open-CV?) algorithm, and while I&#8217;m sure these strategies work against that, our (not very reliable) software looked right through the make-up and noise-glasses.<\/p>\n<p>First test: \u00a0<a href=\"http:\/\/cvdazzle.com\/\" target=\"_blank\">CV Dazzle project.<\/a>\u00a0The artist says the make-up and hairdo effectively stops software from recognizing faces. Let&#8217;s see&#8230;\u00a0we pointed our IP-cam at a computer screen, the easiest way to get images from one system into another.<\/p>\n<p>&nbsp;<\/p>\n<p><img decoding=\"async\" class=\" alignnone\" alt=\"\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-14-1.jpg\" width=\"800\" height=\"734\" \/><\/p>\n<div>\n<dl>\n<dd>CV-Dazzle make-up working as expected. The bottom row of faces is not recognized.<\/dd>\n<\/dl>\n<div><\/div>\n<\/div>\n<p><img decoding=\"async\" class=\"size-full wp-image-1107 alignnone\" alt=\"Netavis-screen-photos-13\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-13-1.jpg\" width=\"800\" height=\"755\" srcset=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-13-1-400x378@2x.jpg 800w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-13-1-400x378.jpg 400w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-13-1-690x651.jpg 690w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><\/p>\n<p>Same here.<\/p>\n<p>&nbsp;<\/p>\n<p><img decoding=\"async\" class=\"size-full wp-image-1106 alignnone\" alt=\"Netavis-screen-photos-12\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-12-1.jpg\" width=\"800\" height=\"533\" srcset=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-12-1-400x267@2x.jpg 800w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-12-1-400x267.jpg 400w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-12-1-690x460.jpg 690w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><\/p>\n<div>\n<dl>\n<dd>Oops. The algorithm got gender and age wrong, but hey. Not too bad!<\/dd>\n<\/dl>\n<div><\/div>\n<\/div>\n<p><img decoding=\"async\" class=\"size-full wp-image-1109 alignnone\" alt=\"Seems to be learning as we go.\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-16-1.jpg\" width=\"800\" height=\"771\" srcset=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-16-1-400x386@2x.jpg 800w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-16-1-400x386.jpg 400w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/02\/Netavis-screen-photos-16-1-690x665.jpg 690w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><\/p>\n<div>\n<dl>\n<dd>It also seems to be learning as we go.<\/dd>\n<\/dl>\n<div><\/div>\n<\/div>\n<p>Next test:\u00a0<a href=\"http:\/\/www.nii.ac.jp\/userimg\/press_details_20121212.pdf\" target=\"_blank\">Isao Echizen&#8217;s modified sun glasses<\/a>:<\/p>\n<p><img decoding=\"async\" class=\"alignnone size-full wp-image-1090\" alt=\"Screen shot 2013-01-23 at 21.54.05\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1.png\" width=\"924\" height=\"718\" srcset=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1.png 924w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1-850x660.png 850w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1-400x311.png 400w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1-690x536.png 690w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Screen-shot-2013-01-23-at-21.54.05-1-400x311@2x.png 800w\" sizes=\"(max-width: 924px) 100vw, 924px\" \/><\/p>\n<p>The project is called &#8220;Privacy Protection Techniques Using Differences in Human and Device Sensitivity&#8221;. If I understand it correctly, infra-red light (invisible to the human eye) in the eye-to-nose bridge region in the face is meant to irritate face recognition algorithms.<\/p>\n<p>Again, we pointed our IP-cam at a computer screen.<\/p>\n<figure id=\"attachment_1101\" aria-describedby=\"caption-attachment-1101\" style=\"width: 800px\" class=\"wp-caption alignnone\"><img decoding=\"async\" class=\"size-full wp-image-1101\" alt=\"Netavis-screen-photos-06\" src=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Netavis-screen-photos-06-1.jpg\" width=\"800\" height=\"533\" srcset=\"https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Netavis-screen-photos-06-1-400x267@2x.jpg 800w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Netavis-screen-photos-06-1-400x267.jpg 400w, https:\/\/sag.khm.de\/wp-content\/uploads\/sites\/15\/2013\/01\/Netavis-screen-photos-06-1-690x460.jpg 690w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><figcaption id=\"caption-attachment-1101\" class=\"wp-caption-text\"><span class=\"Apple-style-span\" style=\"font-size: 13px;line-height: 19px\">\u00a0<\/span><\/figcaption><\/figure>\n<div>\n<dl>\n<dd>Bingo! Even if we take into account that a video of pink-ish light on a computer screen isn&#8217;t the same as actual IR light, it doesn&#8217;t seem to work too well&#8230;<\/dd>\n<\/dl>\n<div><\/div>\n<\/div>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We had some fun at my &#8216;Raumkontrolle&#8217; block seminar, trying out our Netavis Observer &#8216;face recognition&#8217; software on various anti-Anti-Facial Recognition strategies. Seems they all used the same (Open-CV?) algorithm, and while I&#8217;m sure these strategies work against that, our (not very reliable) software looked right through the make-up and noise-glasses. First test: \u00a0CV Dazzle [&hellip;]<\/p>\n","protected":false},"author":15,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_vp_format_video_url":"","_vp_image_focal_point":[],"footnotes":""},"categories":[59,90,52,61],"tags":[],"class_list":["post-1105","post","type-post","status-publish","format-standard","hentry","category-current","category-facial-recognition","category-seminars","category-technology"],"_links":{"self":[{"href":"https:\/\/sag.khm.de\/index.php?rest_route=\/wp\/v2\/posts\/1105","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sag.khm.de\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sag.khm.de\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sag.khm.de\/index.php?rest_route=\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/sag.khm.de\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1105"}],"version-history":[{"count":0,"href":"https:\/\/sag.khm.de\/index.php?rest_route=\/wp\/v2\/posts\/1105\/revisions"}],"wp:attachment":[{"href":"https:\/\/sag.khm.de\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1105"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sag.khm.de\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1105"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sag.khm.de\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1105"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}