Clothes capable of fooling facial recognition systems. This is the leitmotif of the “Manifesto” collection, launched by a young Italian company called Cap_able. The objective of the start-up is to denounce the absence of consent in the use of this technology, indicates Rachele Didero, its co-founder and general manager, interviewed by CNN.
Of the “conflicting algorithms” developed
To trick facial recognition, Cap_able has developed “contradictory algorithms” that generate patterns capable of tricking software. The images are then reconstructed physically, using a computerized and now patented knitting machine. Sweaters, hoodies, pants, T-shirts… the clothes are made in Italy from Egyptian cotton.
Bright colors and original patterns have been developed and tested with YOLO, a facial recognition software widely used around the world. The first results would be encouraging, with a failure of identification in 60 to 90% of cases.
“When I’m in front of a camera, I have no choice whether to give it my data or not. So we create clothes that can give you the opportunity to make that choice. We are not trying to be subversive”insists the entrepreneur.
“Combining fashion and high technology”
The idea came to her while studying at the Fashion Institute of Technology in New York in 2019, after a discussion with an engineer about privacy and human rights. “’The idea of combining fashion and high technology” was born. To continue fooling facial recognition algorithms, the company would need to continue improvements over time alongside technological advancements to counter this modern hacktivism.
To carry out its project, the start-up launched a Kickstarter campaign in 2022 which raised €5,000. The company now plans to integrate the accelerator of the Polytechnic of Milan to refine its business model before presenting itself to investors later this year.
More than a tool in the fight against facial recognition, the “Manifesto” collection is intended as a pretext for “raise awareness of the importance of privacy and human rights by addressing the issue of misuse of facial recognition technology”assures the company on its site.