This site combines two concepts. The first is helping to correct poorly scanned in books. Machines can’t always correctly scan and convert to text images of text. Humans are much better at it. And identifying humans to stop scripts being used say to create free email accounts is often done by using a captcha. You are shown an image and asked to type in the text or numbers. Captcha is short for Completely Automated Public Turing test to tell Computers and Humans Apart .
This site serves captcha images to websites and email anti-spam using text images from badly scanned books. The correct results are then used to update the scanned book image. Quite clever though I do wonder how they know an answer is correct without doing the donkey work in the first place!
Popularity: 11% [?]
Over a year ago, there was a Portent Story on turning photos into 3D at Carnegie Mellon. There is a working website at Stanford University where you can upload a photo (it needs registration) then have it processed into 3D and you can view it using Shockwave.
If you haven’t got Shockwave installed, it will do it for you and once it’s in, you can view your photo or others, zoom in, pan. It’s not bad at all, but it depends upon the original photo and landscapes like the one shown make the best images.
It takes about 5 minutes from uploading until your photo is ready to view. If you’re interested you can download and view the source code in C and C++, but it is a 42 MB file. There’s also a neat integration with Flickr so you can search for photos by tag on Flickr and use those.
Popularity: 15% [?]
Something a little different. You are shown image after image and have to rate them for life likeness, on a scale of 0 to 10 where 10 is most like a face like and 0 isn’t anything like a face.
I suspect this is some variation of a genetic algorithm. This is a way of trying to solve a problem that doesn’t lend itself to normal computation by using a process akin to genetic selection. Interesting.
Popularity: 13% [?]