{"id":2869,"date":"2016-03-23T21:29:44","date_gmt":"2016-03-24T01:29:44","guid":{"rendered":"http:\/\/subjugator.org\/?p=2869"},"modified":"2016-07-24T21:30:38","modified_gmt":"2016-07-25T01:30:38","slug":"begining-stages-of-machine-learning","status":"publish","type":"post","link":"http:\/\/subjugator.org\/?p=2869","title":{"rendered":"Begining Stages of Machine Learning"},"content":{"rendered":"<div>\n<p><u>This Week We<\/u><\/p>\n<div><\/div>\n<\/div>\n<div>\n<ul>\n<li><span style=\"line-height: 1.5;\">Made a tool for segmenting training data<\/span><\/li>\n<li>Experimented with using an SVM for segmentation<\/li>\n<li>Worked on stuff for easily segmenting\/labelling planes in PCL\n<ul>\n<li>+ distinguishing things &#8220;on top&#8221; of the plane from the plane itself<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/div>\n<div>\n<div>The training tool: The plot on the right is HSV-&gt;XYZ. You draw a box around the desired color groups, and use the resulting segmentation.<\/div>\n<div>\n<p>This is\u00a0less annoying than manually drawing on images.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\" wp-image-2872 aligncenter\" src=\"http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/easy_labeller-300x190.png\" alt=\"easy_labeller\" width=\"399\" height=\"253\" srcset=\"http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/easy_labeller-300x190.png 300w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/easy_labeller-768x487.png 768w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/easy_labeller-1024x649.png 1024w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/easy_labeller.png 1618w\" sizes=\"(max-width: 399px) 100vw, 399px\" \/><\/p>\n<\/div>\n<div><\/div>\n<div><\/div>\n<div>A\u00a0SVM was trained on the segmentation data\u00a0and it ran\u00a0in ~near~ real-time on a super downsampled input image.<\/div>\n<div>This is one frame of the SVM running on the video. It was just a 2-second demo, the threshold segmentation looks a lot better.<\/div>\n<div>Sci-kit learn&#8217;s svm implementation struggles to fit at more than 10k samples, which is another\u00a0problem.<\/div>\n<\/div>\n<div><img loading=\"lazy\" decoding=\"async\" class=\" wp-image-2871 aligncenter\" src=\"http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/svm_trainer-300x153.png\" alt=\"svm_trainer\" width=\"400\" height=\"204\" srcset=\"http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/svm_trainer-300x153.png 300w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/svm_trainer-768x391.png 768w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/svm_trainer-1024x521.png 1024w, http:\/\/subjugator.org\/wp-content\/uploads\/2016\/07\/svm_trainer.png 1668w\" sizes=\"(max-width: 400px) 100vw, 400px\" \/><\/div>\n","protected":false},"excerpt":{"rendered":"<p>This Week We Made a tool for segmenting training data Experimented with using an SVM for segmentation Worked on stuff for easily segmenting\/labelling planes in PCL + distinguishing things &#8220;on top&#8221; of the plane from the plane itself The training tool: The plot on the right is HSV-&gt;XYZ. You draw a box around the desired [&hellip;]<\/p>\n","protected":false},"author":23,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[4],"tags":[],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/posts\/2869"}],"collection":[{"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/users\/23"}],"replies":[{"embeddable":true,"href":"http:\/\/subjugator.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2869"}],"version-history":[{"count":3,"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/posts\/2869\/revisions"}],"predecessor-version":[{"id":2875,"href":"http:\/\/subjugator.org\/index.php?rest_route=\/wp\/v2\/posts\/2869\/revisions\/2875"}],"wp:attachment":[{"href":"http:\/\/subjugator.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2869"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/subjugator.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2869"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/subjugator.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2869"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}