Skip to Content

Discovered at DEFCON 27: automated license plate readers (ALPRs) being hoodwinked by clothing

Discovered at DEFCON 27: automated license plate readers (ALPRs) being hoodwinked by clothing

It seems Joe Public is shouting “privacy here, privacy there, privacy everywhere,” as people are pushing back against certain technologies that could, or people believe could, misidentify them and track, monitor and record their actions, or be the catalyst to their personal information and identity being stolen.

It's a double-edged sword really; people want to use the technology to ensure safety and security, but at the same time, they want no interference with their privacy. It's all or nothing. Unfortunately, we aren't at a point with technology where “good” people are automatically excluded from the “bad.” However, one solution to protect privacy presented itself about a week ago at none other than DEFCON 27.

As over 25,000 security professionals and researchers, federal government employees, lawyers, journalists, and of course, hackers with an interest in anything and everything that can be hacked descended on Las Vegas' Paris, Bally's, Flamingo and Planet Hollywood Convention Centers, professional ethical hacker and now, fashion designer, Kate Rose, debuted her weapon of choice against ALPRs and surveillance — t-shirts, hoodies, jackets, dresses and skirts.

Knows as Adversarial Fashion, each garment is purposely designed to trigger ALPRs and inject data rubbish into systems used by states and its contractors, believed by some to monitor and track civilians. Rose tested a series of modified license plate images with commercial ALPR APIs and created fabric patterns that read into LPRs as if they are authentic license plates. Priced at no more than 50 bucks, tops, you too can now fool ALPRs with your clothes!

Don't feel like shelling out your hard-earned money? Not to worry! Rose lists all the resources needed to make your own computer vision-triggering fashion and fabric designs on her site, along with a hyperlinked list of libraries and APIs, image editing tools, color palette extraction tools and textile pattern tutorials. In addition, slides from her DEFCON 27 Crypto and Privacy Village talk, “Sartorial Hacking to Combat Surveillance,” offering the following how-to guide of designing your own anti-surveillance clothes:

  1. Choose a recognition system and experiment with design constraints, starting with high confidence images.
  2. Test tolerances by making slight modifications to source images.
  3. Make notes of “cue” attributes that affect confidence scores.
  4. Plot enough images to determine what seems to work.
  5. Use images that work to design a pattern and digitally print it onto fabric.

I'm not too sure if this is a 5-step method to early retirement, but I can say people are demanding privacy and obviously, being very creative in their fight for it.

Comments

To comment on this post, please log in to your account or set up an account now.