What will replace face recognition cameras
2023.01.16 07:33
What will replace face recognition cameras
By Kristina Sobol
Budrigannews.com – The redheaded man approaches the camera while sporting what appears to be the ultimate Christmas sweater. He is surrounded by a yellow quadrant. The man is immediately recognized by facial recognition software as… a giraffe?
This case of mistaken identity was planned in advance, not by accident. The sweater is part of the Italian startup Cap_able’s debut Manifesto collection. It includes dresses, hoodies, pants, and tops in addition to tops. An “adversarial patch” is a pattern on each one created by artificial intelligence algorithms to frighten facial recognition software: Either the wearer is not recognized by the cameras, or they mistake them for a dog, a giraffe, a zebra, or one of the other animals incorporated into the pattern.
Rachele Didero, co-founder and CEO, asserts, “When I’m in front of a camera, I don’t have a choice of whether I give it my data or not.” As a result, we are developing clothing that will enable you to make this choice. We are not attempting anything subversive.
Didero, 29, who is pursuing a PhD in “Textile and Machine Learning for Privacy” at Politecnico di Milano and has also worked at MIT’s Media Lab, claims that she came up with the idea for Cap_able while she was on a Masters exchange at the Fashion Institute of Technology in New York. She read about tenants in Brooklyn who had resisted their landlord’s plans to install a facial recognition entry system in their building while she was there.
She states, “This was the first time I heard about facial recognition.” Because one of my friends was an engineer in computer science, we both thought, “This is a problem; maybe we can combine fashion design and computer science to create something you can wear every day to protect your data.”
The simple part was coming up with the concept. They had to find and then design the right “adversarial algorithms” to help them create images that would fool facial recognition software before they could make it a reality. Either they would use the algorithm to modify the image, say, of our giraffe, or they would create it. Or, they gave the algorithm the colors, dimensions, and shape they wanted for the pattern or image.
Didero says, “You need a mindset between engineering and fashion.”
They had to test the images on a well-known object detection system called YOLO, which is one of the algorithms in facial recognition software that is used the most frequently.
Using a Computerized Knitwear Machine, which resembles a loom and a huge barbecue, they would then create a physical version of the pattern in a process that is now protected by a patent. They could then create their range, all made in Italy from Egyptian cotton, with a few minor adjustments to achieve the desired appearance, size, and position of the images on the garment.
According to Didero, when tested with YOLO, the current clothing items function 60% to 90% of the time. While Cap_able’s adversarial algorithms will get better, the software it is attempting to deceive may also get better, if not faster.
Brent Mittelstadt, associate professor and director of research at the Oxford Internet Institute, describes the situation as “an arms race.” He compares it to a conflict between software that detects deep fakes and software that makes them. Except for clothing, updates cannot be downloaded.
He stated, “It may be that you purchase it, and then it is only good for a year, or two years, or five years, or whatever it takes to actually improve the system to the point where it would ignore the method used to fool them in the first place.”
He also says that these clothes might end up being just a niche product because they start at $300.
However, their impact might extend beyond safeguarding the privacy of those who purchase and wear them.
According to Woodrow Hartzog, a professor at Boston University School of Law, “one of the key advantages is it helps create a stigma around surveillance, which is really important to encourage lawmakers to create meaningful rules, so the public can more intuitively resist really corrosive and dangerous types of surveillance.”
Cap_able is not the first project to combine design and privacy protection. At the most recent World Cup in Qatar, the creative agency Virtue Worldwide created flag-themed face paint for fans who wanted to fool the numerous facial recognition cameras in the emirate.
Adam Harvey is a Berlin-based artist who works with data, privacy, surveillance, and computer vision. He has made makeup, clothing, and apps that help people feel more secure. He created Hyperface in 2016, a textile with “false-face computer vision camouflage patterns” that may be considered an artistic precursor to Cap_able’s current commercial endeavors.
Shira Rivnai Bahir, a lecturer at Israel’s Reichman University’s Data, Government, and Democracy program, states, “It’s a fight, and the most important aspect is that this fight is not over.” Even if it doesn’t completely protect us, going to street protests gives us more confidence or the impression that we aren’t completely giving ourselves over to the cameras.
The Hong Kong protesters’ use of umbrellas, masks, and lasers as some of the more analog ways people have fought back against the rise of the machines is cited by Rivnai Bahir, who is about to submit her PhD thesis on the role of anonymity and secrecy practices in digital activism. However, the authorities are able to easily identify these and seize them. It might be more difficult to follow someone else’s pattern for a sweater.
Late last year, Cap_able launched a Kickstarter campaign. It brought in €5,000. Before presenting its business plan to investors later this year, the company now intends to enroll in the Politecnico accelerator program.
Didero says that when she wears the clothes, people say they think they are “cool” before she admits: Maybe it’s because I live in New York or Milan, where it’s not as crazy!
Thankfully, more subdued ranges are on the way, with patterns that aren’t as obvious to the naked eye but can still confuse cameras. In places like China, where facial recognition was a key part of efforts to identify Uyghurs in the northwestern region of Xinjiang, or Iran, which is reportedly planning to use it to identify hijab-less women on the metro, it may also help Capable-clothed individuals avoid sanction from the authorities.
Although Big Brother’s eyes may become increasingly omnipresent, it’s possible that in the future he will instead see giraffes and zebras.
