Scientists from the University of Maryland have invented a sweater that can hide the wearer from facial recognition systems. New research shows that it is possible to create clothing that can block facial recognition with cameras placed in public places.Experts have produced a confusing picture for AI-based facial recognition systems. The image in question resembles a photo of people at the vegetable market from behind the counter, but in reality we are talking about a completely synthetic image.
The YOLOv2 algorithm has been tricked
This particular pattern created was applied to a sweater. As a result, the person wearing this sweater could not be detected by artificial intelligence-based facial recognition systems. Let’s not forget that the tests are carried out on the YOLOv2 detector, which can detect real-time objects.
The researchers describe their new invention as “This stylish sweater is perfect for keeping you warm in the winter at the office or on the road. Its waterproof micro fleece lining, modern cut and anti-AI patterns help you hide from object detectors”.
52809
