{"id":111918,"date":"2018-08-29T16:14:12","date_gmt":"2018-08-29T23:14:12","guid":{"rendered":"http:\/\/erickimphotography.com\/blog\/?p=111918"},"modified":"2018-09-26T19:20:18","modified_gmt":"2018-09-27T02:20:18","slug":"how-photographers-can-see-like-machines","status":"publish","type":"post","link":"https:\/\/erickimphotography.com\/blog\/2018\/08\/29\/how-photographers-can-see-like-machines\/","title":{"rendered":"How to See Like a Machine"},"content":{"rendered":"<p>We are in a <a href=\"http:\/\/erickimphotography.com\/blog\/2017\/06\/17\/brave-new-world-of-photography\/\">brave new world of photography<\/a>; we can leverage machine learning and <a href=\"http:\/\/erickimphotography.com\/blog\/2018\/08\/12\/why-ai-artificial-intelligence-is-good-for-photographers\/\">artificial intelligence<\/a> to make better photos! Some things I&#8217;ve been learning:<\/p>\n<p><!--more--><\/p>\n<h2>DOWNLOAD<\/h2>\n<p>PDF:<\/p>\n<blockquote><p>\n  <a href=\"http:\/\/erickimphotography.com\/blog\/wp-content\/uploads\/2018\/08\/Machine-Learning-for-Photographers.pdf\">Machine Learning for Photographers<\/a>\n<\/p><\/blockquote>\n<p>Also download photo examples:<\/p>\n<blockquote><p>\n  <a href=\"https:\/\/www.dropbox.com\/s\/ubv880r2ywwm86y\/Machine%20Learning%20for%20Photographers%20-%20Photo%20Examples.zip?dl=0\">ZIP: Photography Picture Examples<\/a>\n<\/p><\/blockquote>\n<hr \/>\n<h2>Behind the Scenes Video:<\/h2>\n<p>The making of video, with screen capture and Camtasia on my laptop:<\/p>\n<p><iframe loading=\"lazy\" title=\"How I Work: Machine Learning for Photographers\" width=\"500\" height=\"375\" src=\"https:\/\/www.youtube.com\/embed\/GDfkv1qY9bc?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<p><a href=\"http:\/\/www.youtube.com\/watch?v=GDfkv1qY9bc\">Watch on YouTube &gt;<\/a><\/p>\n<hr \/>\n<h2>Image Segmentation<\/h2>\n<p>To better study and look at images, segment the images (something I learned from machine learning):<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-28-at-10-14-16-pm.png\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>How do machines see images?<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-28-at-10-14-34-pm.png\" alt=\"\" \/><\/figure>\n<p>To start off, I want to state that humans are far superior to machines in almost every way. Let us not get fooled by the &#8216;golden calf&#8217; of AI (artificial intelligence):<\/p>\n<blockquote><p>The purpose is for us to leverage machines to make better humans.<\/p><\/blockquote>\n<p>Not,<\/p>\n<blockquote><p>Trying to use humans to make better machines.<\/p><\/blockquote>\n<p>To re-iterate:<\/p>\n<blockquote><p>Let us leverage machine learning to better improve ourselves.<\/p><\/blockquote>\n<p>So I don&#8217;t think the purpose is for humans to see like machines (remember Arnold S as the Terminator, when he sees the world in black-and-red). What I propose is that we study how machines see images, as a way to guide how we can better see as humans.<\/p>\n<p>Or in other words:<\/p>\n<blockquote><p>Let us study machine learning, to better understand how we (as humans) think and see!<\/p><\/blockquote>\n<hr \/>\n<h2>Simplify<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-28-at-10-14-45-pm.png\" alt=\"\" \/><\/figure>\n<p>The interesting thing about how machines see images:<\/p>\n<blockquote><p>They simplify it.<\/p><\/blockquote>\n<p>When I study my own black and white photos, I believe my best photos are my simple photos. Even if you study great artists like Picasso and Matisse, they always strived for elegant images&#8211; simple and elegant. Think of Matisse&#8217;s cutouts &#8212; he was able to abstract the human body into the simplest shape, form, and color.<\/p>\n<hr \/>\n<h2>Figure-to-ground<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/eric-kim-abstract-photography1-copy.jpg\" alt=\"\" \/><\/figure>\n<p>Figure (the subject) and ground (background): you want maximal contrast between the two elements.<\/p>\n<p>One thing we can use: &#8220;Gaussian Blur&#8221; in Photoshop, to abstract the images, to just see the position of our visual elements, and to simplify the contrast:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/1-9.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/2-9.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/3-8.jpg\" alt=\"\" \/><\/figure>\n<h2>Does the photo work well as a small thumbnail?<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/eric-kim-abstract-photography11.jpg\" alt=\"\" \/><\/figure>\n<p>To better understand whether your compositions are strong or not&#8211; look at them as small thumbnails. If your photos work as small thumbnails, it probably means your composition is strong!<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/4-5.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/5-4.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/6-4.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/7-2.jpg\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>Bounding Boxes<\/h2>\n<p>There is already artificial intelligence that can automatically put &#8216;bounding boxes&#8217; (little red boxes) around subjects. This is how self-driving cars see pedestrians (not to run them over):<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-15-25-am.png\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-15-38-am.png\" alt=\"\" \/><\/figure>\n<p>Even more insane&#8211; machines can automatically color, segment, and classify different subjects in a scene (real time):<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-18-29-am1.png\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-19-13-am.png\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-19-22-am.png\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-19-30-am.png\" alt=\"\" \/><\/figure>\n<p>A simple way we (humans) can look at photographs:<\/p>\n<blockquote><p>Let us put little red boxes around different visual elements in photos.<\/p><\/blockquote>\n<h2>The Cherry on Top<\/h2>\n<p>For example, I like to look at the &#8216;cherry on top&#8217; for subjects in the photograph&#8211; small details which I find interesting. For example, in this photograph, this little boy in the background is a detail I love:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/bounding-box.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/bounding-box-2.jpg\" alt=\"\" \/><\/figure>\n<p>But this is the difficult thing&#8211; most people (when they see the picture really small, perhaps on a phone), they won&#8217;t see the detail of the boy in the background:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/bounding-box-3.jpg\" alt=\"\" \/><\/figure>\n<p>This is why I like printing out photos big&#8211; so you can actually see small details!<\/p>\n<p>And to be honest, 99% of your viewers won&#8217;t notice these &#8216;cherry on top&#8217; details in the photo. But it doesn&#8217;t matter&#8211; as long as <em>YOU<\/em> see the detail and like it&#8211; that is enough!<\/p>\n<hr \/>\n<h2>Color Segmentation<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/red-1.jpg\" alt=\"\" \/><\/figure>\n<p>We can also apply gaussian blur to color photos, to better understand the color palette of our photos:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/red-2.jpg\" alt=\"\" \/><\/figure>\n<p>I used the color picker in Photoshop (hotkey &#8220;I&#8221; for &#8220;I[eye] dropper&#8221;). And painted these colors:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/red-3.jpg\" alt=\"\" \/><\/figure>\n<p>If I make the background grey, you can see the colors:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/red-4.jpg\" alt=\"\" \/><\/figure>\n<p>So you can see the basic colors: rogue red, pale yellow, and green in the bottom-left.<\/p>\n<hr \/>\n<h2>Keep the colors simple<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-1.jpg\" alt=\"\" \/><\/figure>\n<p>For color photography, I generally think the simpler, the better.<\/p>\n<p>For example this picture of this woman, the primary color is pink (with Gaussian Blur applied):<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-2.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-3.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-4.jpg\" alt=\"\" \/><\/figure>\n<p>The final with the background subtracted:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-5.jpg\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>Which details do you look at in a photo?<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-6.jpg\" alt=\"\" \/><\/figure>\n<p>If you use the concept of a &#8216;bounding box&#8217; &#8212; you can use it as a tool to identify things you find interesting in a scene&#8211; like the woman smoking a cigarette, and two Red Bulls in her left hand.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/pink-7.jpg\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>Feature Map<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-3-38-36-pm.png\" alt=\"\" \/><\/figure>\n<p>When you are shooting photos, try to abstract your vision &#8212; simplify it.<\/p>\n<p>For example, this is why I recommend shooting high-contrast black and white mode, because it helps you <em>see<\/em> better when you&#8217;re taking photos.<\/p>\n<p>For example on a RICOH GR II, I recommend using the &#8216;high contrast black and white mode&#8217; in JPEG, and when you&#8217;re composing your photos, keep the photos simple.<\/p>\n<h2>Silhouette<\/h2>\n<figure><img decoding=\"async\" src=\"\/\/\/Users\/erickim\/Library\/Mobile%20Documents\/27N4MQEA55~pro~writer\/Documents\/silhouette%201.jpg?md5=63e3ffab7a84b785d14ccaaea977d614\" alt=\"\" \/><\/figure>\n<p>This is how silhouettes work &#8212; we create a &#8216;feature map&#8217; (the gradients between black, grey, and white.<\/p>\n<hr \/>\n<h2>Edges<\/h2>\n<p>Also another way to look at photos; try to simplify the edges, to see the textures in the photo:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette-2.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette-3.jpg\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>Silhouette of Faces<\/h2>\n<figure><img decoding=\"async\" src=\"\/\/\/Users\/erickim\/Library\/Mobile%20Documents\/27N4MQEA55~pro~writer\/Documents\/silhouette%204.jpg?md5=63e3ffab7a84b785d14ccaaea977d614\" alt=\"\" \/><\/figure>\n<p>Another idea:<\/p>\n<p>As humans, we intuitively know the proportions of a human face, and also the silhouette of faces.<\/p>\n<p>To see better, let us apply maximal contrast to the photo:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette-5.jpg\" alt=\"\" \/><\/figure>\n<p>Then, a cyan cutout of the woman:<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette-6.jpg\" alt=\"\" \/><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/silhouette-7.jpg\" alt=\"\" \/><\/figure>\n<hr \/>\n<h2>Conclusion: The Future of Photography<\/h2>\n<figure><img decoding=\"async\" src=\"https:\/\/ekizz.files.wordpress.com\/2018\/08\/screen-shot-2018-08-29-at-9-20-01-am.png\" alt=\"\" \/><\/figure>\n<p>I&#8217;m currently doing more studying on Augmented Reality (AR), especially the &#8216;AR CORE&#8217; platform from Google, as well as their &#8220;Tensor Flow&#8221; platform. The basic idea I have is this:<\/p>\n<blockquote><p>One day, while you&#8217;re taking photos on your photos, you will get compositional grids, and &#8216;bounding boxes&#8217; to help assist your compositions.<\/p><\/blockquote>\n<p>I already have this functionality on &#8220;intelligent Auto&#8221; mode on my Lumix G9&#8211; when I am composing a scene, the LCD screen will automatically put a yellow box around my subject in the frame.<\/p>\n<h3>Augmented Reality for Photographers<\/h3>\n<p>I think the future of photography is this:<\/p>\n<blockquote><p>Photographers will leverage artificial intelligence (AI) and machine learning to take better photos (improved composition, or awareness of subjects).<\/p><\/blockquote>\n<hr \/>\n<h2>The future is already here<\/h2>\n<p>For further reading, I recommend:<\/p>\n<ol>\n<li>Research &#8220;AR CORE&#8221; from Google<\/li>\n<li>Research &#8216;computational photography&#8217;<\/li>\n<li>Check out the Google AI Experiments<\/li>\n<li>Download some &#8220;AR&#8221; (augmented reality) apps on your phone or iPad and play around with it<\/li>\n<li>Think about how photographers, individuals, and artists can leverage virtual reality, augmented reality, and machine learning to make better art!<\/li>\n<\/ol>\n<p>ERIC<\/p>\n","protected":false},"excerpt":{"rendered":"<p>We are in a brave new world of photography; we can leverage machine learning and artificial intelligence to make better photos! Some things I&#8217;ve been learning:<\/p>\n","protected":false},"author":1,"featured_media":111919,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"advanced_seo_description":"","jetpack_seo_html_title":"","jetpack_seo_noindex":false,"jetpack_seo_schema_type":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_post_was_ever_published":false},"categories":[17],"tags":[],"class_list":["post-111918","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-posts"],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/erickimphotography.com\/blog\/wp-content\/uploads\/2018\/08\/Screen-Shot-2018-08-28-at-10.14.16-PM.png?fit=1274%2C822&ssl=1","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/posts\/111918","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/comments?post=111918"}],"version-history":[{"count":5,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/posts\/111918\/revisions"}],"predecessor-version":[{"id":114170,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/posts\/111918\/revisions\/114170"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/media\/111919"}],"wp:attachment":[{"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/media?parent=111918"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/categories?post=111918"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/erickimphotography.com\/blog\/wp-json\/wp\/v2\/tags?post=111918"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}