<body><script type="text/javascript"> function setAttributeOnload(object, attribute, val) { if(window.addEventListener) { window.addEventListener('load', function(){ object[attribute] = val; }, false); } else { window.attachEvent('onload', function(){ object[attribute] = val; }); } } </script> <div id="navbar-iframe-container"></div> <script type="text/javascript" src="https://apis.google.com/js/platform.js"></script> <script type="text/javascript"> gapi.load("gapi.iframes:gapi.iframes.style.bubble", function() { if (gapi.iframes && gapi.iframes.getContext) { gapi.iframes.getContext().openChild({ url: 'https://www.blogger.com/navbar.g?targetBlogID\x3d3323185\x26blogName\x3dbarnskiblog\x26publishMode\x3dPUBLISH_MODE_BLOGSPOT\x26navbarType\x3dBLUE\x26layoutType\x3dCLASSIC\x26searchRoot\x3dhttps://barnski.blogspot.com/search\x26blogLocale\x3den_US\x26v\x3d2\x26homepageUrl\x3dhttp://barnski.blogspot.com/\x26vt\x3d5885749210978744567', where: document.getElementById("navbar-iframe-container"), id: "navbar-iframe", messageHandlersFilter: gapi.iframes.CROSS_ORIGIN_IFRAMES_FILTER, messageHandlers: { 'blogger-ping': function() {} } }); } }); </script>

barnskiblog

Barney's blog. Just a load of old shite really.


Peekaboom: Play this game and teach computers to see.
Basically, computers are really good at maths, which can be used to represent and manipulate all kinds of stuff, including your music, photos, videos, documents and so on. The problem is, computers are pretty crap at interpreting that stuff the way we do, so whilst they can change tones and colours in a photo for example, they can't actually identify what the photo is of. In fact, computers generally can't even scan images for the tell-tale signs that we use to differentiate between people and objects, so whilst a computer can manipulate an image as instructed, it can't tell if the picture is of Winston Churchill or a Land Rover.
This is particularly relevant if you've ever spent time adding metadata to your photos - I have spent hours adding keywords to iPhoto, and still have a way to go before my photo library is completely up to date. This is really useful though, as once I have added a keyword for "beach" to all the pictures that are of a beach then I can instantly find all those pictures using that keyword. Even better, Spotlight reads those keywords, so if I do a Spotlight search for "Mum", then my results include not only e-mails and documents but also all the photos of my Mum in my library (or at least, all the ones I have tagged with the appropriate keyword).
The point is that I have to do the work of adding the keyword metadata to the photos myself, as my computer is not capable of understanding what the subject or content of a picture is. That's where Peekaboom comes in - it's an online picture guessing game created at Carnegie Mellon University. The cool thing is, all the guesses and data you supply whilst playing is used as training data for algorithms that are being developed to address the problem I just described. These algorithms will hopefully develop into a form that can be implemented in the future to help computers identify what is in a picture.
I like stuff like this :)

read more | digg story
« Home | Next »
| Next »
| Next »
| Next »
| Next »
| Next »
| Next »
| Next »
| Next »
| Next »