A brief post this evening, since I’m a bit worn out. Today I was teaching my introductory programming course, and we happened to be talking about libraries so I demonstrated a little of the OpenCV library you can use with Processing. Of particular cool-factor is the fact that it gives you some pretty accurate and easy face detection – I was mostly using it as a “libraries give you superpowers” affair for that very reason.
But I found myself wanting to “play” with the facial detection algorithm. Not so much to work out how it works in any specific way, but more to see what it reckons a face is, and how you can interact with that recognition. So I started drawing faces in yellow chalk on the blackboard, and it was weirdly fun trying to draw a face that the program would recognise and draw a red rectangle around. (Click to enlarge the image if you particularly want to see what was on my screen while doing this.)
Now, in this scenario there’s me standing at a blackboard in front of a camera, and I’m drawing in yellow chalk on the board, and the camera is being used by face detection software, which is displaying the camera feed of my drawings (in real time) and drawing rectangles around what it thinks are faces. This feels like a strangely fulfilling interaction with a computer, somewhat at the level of code, and somewhat not. There was a sense in which it felt like I was “coding”, except with drawing. By drawing different faces I was basically interacting more or less directly with a facial recognition algorithm. And that felt cool and good.
Which relates a bit to some of the discussion we had around Henning Eichberg’s talk today at ITU, particularly this notion that computers as systems encourage all kinds of rules and “this is the way it is”, and bring with that the temptation to feel that creating video games that are sets of rules is desirable. And so the obvious corollary, how would be avert this if we wanted to? It does seem rather limiting to notions of play to lean heavily of computers as devices for reproducing rules.
I don’t have an answer or even a philosophical point, but I found it interesting to have a highly playful and fun interaction with something as serious and computer sciencey as a facial recognition algorithm, via the act of drawing on a blackboard. Seems like there’s a pleasing step there away from computers as enforcing rules on us, and computers as providing a field of expression and play.
Make of it what you will.