First, let’s note my word creativity in the title. While nothing is original, I don’t recall seeing this before. Google tells us this is a relatively unused word (thankfully?).

Second, the reason I post is a now-old entry on Engadget. The post is short, so I’ll quote it in its entirety:

Instead of standing around using the same old channels to affect change or whining about what a violation of civil liberties it is that Austrian police can install closed-circuit surveillance cameras in public spaces without a court order, activist group Quintessenz took it to the streets to fight back, both high tech and low. Before moving on to more advanced methods, Quintessenz’s early efforts merely blinded the cameras with lasers and balloons, but they’ve since upgraded to techniques that scramble camera signals and even intercept their video feeds and automatically place Dirty Deeds Done Dirt Cheap-style black stripes over peoples’ eyes. Way to subvert, boys and girls — ain’t no fire worth stoking like the Big Brother civil liberties violations fire.

It’s an interesting exercise and part of the larger trend toward integrating technology into public art and activism (for more examples, see Wired article on tech graffiti, for which I can not find a link).

However, the more effective way of reducing the effectiveness of recognition by video camera (by human or software) is by covering the lower half of your face, not your eyes. About two years ago, a Carnegie-Melon and University of Pittsburgh research team used video of me moving around to demonstrate facial-recognition software. The software was primarily being developed to measure face muscle movement, with the intent of doing face reconstructions, I believe, but had a number of spinoff applications including eventually being used as face-recognition software.

I met with the team because they wanted my permission, which I gave, to use the video of me. The primary investigator for the team, whose name I can not recall, was generous with his time, answered my many questions, and even demonstrated the software. One of the questions I asked was what is the most effective way of stumping face-recognition software, such as the type used by police forces. He said even the best packages had tremendous difficulty matching faces that had their bottom half covered (e.g., faces covered with a bandanna around the nose and mouth). This, he said, is much more effective than covering your eyes (e.g., using sunglasses) or top of your head (e.g., using a hat).