The presence of cameras in our streets is increasingly common, but perhaps more important is the progress in the development of algorithms capable of automating surveillance tasks. For example, programs that are capable of detecting people "suspicious" by their appearance, behavior or because they appear in a database.
At the same time that this technology expands in our cities, there are growing movements that seek to counteract it; although perhaps the last project born to avoid the look of the cameras is the most curious.
The hacker and designer Kate Rose has presented at DefCon, one of the most important cybersecurity conferences of the year, a clothing line designed to avoid cameras and their algorithms.
Clothes to deceive cameras
Specifically, Rose has focused on ALPR algorithms, automatic license plate readers, to confuse cameras and make them believe they are recording a car, and not a real person.
That is why the designs of these dresses are mainly based on false license plates; some designs in fact consist solely of those license plates on a single-colored background, while others include numbers in the middle of a circuit pattern, or with text taken from the fourth amendment of the US constitution, which protects against registrations. and arbitrary detentions.
These designs take advantage of the limitations of ALPR, and how they usually have a very wide field of view, but have problems with "specificity"; In other words, if they detect certain components, such as a license plate, they immediately assume that they are recording a vehicle.
These patterns have been designed so that they can be easily read by cameras in good lighting situations; Although its creator recommends taking a size larger than we have, so that the patterns are as flat as possible and the deception works better.
At the moment this clothing line is only conceptual, a protest rather than a real attempt to dodge cameras; Although in theory they work, you could not trust it much if you really want to be "invisible" to the cameras.