Computer Vision

How do you program a computer to understand what's in a picture? How do you programatically define the difference between my face and yours, an A and a B, a cat and a dog or a gun and a laptop.

As humans, it's hard to understand that there's a problem at all - cat's and dogs are just different. What is the essence of that difference though? What set of rules define what makes a dog-image doggy and a cat-image catty?

Detecting objects in images is the problem of recognition and is a part of the field of computer vision, a subject I researched at Oxford Medical Vision Lab for two years after my first degree.