Facial recognition technology is becoming more common. Some startups are struggling with this
The new collection CAP_able is designed toto confuse facial recognition algorithms. Manifesto's debut line includes hoodies, tops and dresses with a special pattern that looks like an abstract print from a distance. But it is actually modeled after the coloring of various animals, including giraffes, zebras and dogs. This is how the Italian designer tried to deceive the smart camera system. She will not be able to identify the owner or will mistake him for an animal.
The idea is to deceive the software forfacial recognition came to the girl's mind when she was studying for a master's degree in exchange at the Fashion Institute of Technology in New York. The designer read about how Brooklyn residents opposed the homeowner's plans to install such a system.
Using a popular object detection systemYOLO, as well as computerized knitting machines from Italy, Didero and her startup colleagues figured out how to optimally place images on Egyptian cotton materials to create truly unique garments. They fooled algorithms like YOLO about 60-90% of the time. In the future, Didero plans to create patterns that are less noticeable to the human eye, but can fool facial recognition systems.
CAP_able launched a campaign forKickstarter at the end of last year and raised about 5,000 euros. To improve the business model, they plan to join the Politecnico accelerator program. Now the cost of one piece of clothing starts from $300.
Read more:
Land in China drilled to a record depth
It's not about the Earth: scientists explained why the solar system is the rarest
Medieval fortress accidentally discovered in the forest: the find surprised scientists
Cover photo: The Manifesto Collection, CAP_able