The Pros and Cons of Facial Recognition Software
The facial recognition use case debates are still controversial, as the technology can benefit society but could also be used to wreak havoc on people’s privacy.
While all this seems like facial recognition is a technology that could improve our safety, the reality is far from perfect. In fact, a technology that can track down someone or keep a unique faceprint of an individual in a database without their consent is threatening and could take us one step closer to the terrifying Orwellian future.
Facial recognition software relies on machine learning and AI algorithm to detect faces from photos or videos. These algorithms are usually taught to find the eyes first and then map the entire face by adding other elements, such as eyebrows, a nose, a mouth, teeth, etc.
Photo illustration: Freepik
Nevertheless, biometric software such as facial recognition relies on algorithms, which could be biased. For example, AI facial recognition software isn’t always equipped to recognize or even differentiate people of all skin colors in the same way. Despite that, facial recognition is one of the most widespread applications of AI, and people use it everywhere, from health and safety to security departments.
Pros of facial recognition:
- Helps locate missing people: Integrating facial recognition software with traffic cameras could allow law enforcement to find missing persons faster. It would broaden the search and cover areas yet to be scouted by local authorities. It could also be an effective tool for tracking suspects and people with criminal records.
- Improves security: Facial recognition access control systems (software for controlling access to properties or specific building parts) are used in many industries, ranging from healthcare to real estate.
- Improves efficiency: Smartphone manufacturers, for example, offer facial recognition authentication, allowing users to access the phone or a specific program or app without a password. Although facial recognition as an authentication method could improve security, the technology needs to be more advanced and reliable enough to replace good old passwords.
Cons of facial recognition:
- Violates privacy and fundamental rights: The idea of using facial recognition and merging it with security and traffic cameras isn’t attractive to most people, as it would conflict with our fundamental rights, including the right to freedom and personal identity.
- Could be misused: Facial recognition software itself could have unlawful applications. People could use it to stalk their victims, extort money, and commit identity theft or other crimes. Besides that, faceprints are stored in breachable databases. Obtaining such datasets is a violation of privacy and, in some cases, safety.
- Produces errors: Human errors happen, but robots and AI bots make mistakes, too. Some would even argue they have more errors than humans. Nevertheless, what matters here is that AI makes mistakes. If we start using facial recognition that produces errors in identifying criminals or suspects on the street, this approach may implicate innocent individuals.
- AI Bias: From ChatGPT to facial recognition software, there’s no doubt that AI is biased. Biased or incomplete information used to train bots produces limited AI output. Until we find a way to fix this issue, ethnic groups and minorities will continue to experience bias through technology.
Cap_able Is Tricking Facial Recognitions With Its Stylish Patterns
Cap_able is an Italian fashion design studio founded in 2021 with the mission to help people alleviate discomfort regarding facial recognition software and its data collection. One of their primary goals is to find a link between design and technology and open the discussion about current issues such as data collection, privacy, and violation of human rights.
The founder and CEO of Cap_able is Rachele Didero, who also created a fabric that confuses facial recognition software. Rachele Didero won the ADI Compasso d’Oro Award for her patents, and her project is continually growing and getting deserved recognition from prominent players in the fashion industry.
Source: Capable Design
Cap_able produces knitting textiles using specific knitting machines, gauges, and yarn, resulting in fabulous knitwear collections.
It says on the Cap_able official website that “knitwear gives texture and wearability as well as makes the design more durable than a printed one. Knitwear is practical and comfortable for those who move around the city and for those who travel for pleasure or for work. Among the other benefits, the garments don’t need to be ironed.”
Cap_able is a proud creator of dresses, shirts, trousers, and other streetwear clothing. So far, the team behind the project has released one collection called Manifesto. Each garment in this collection is not recognizable by facial recognition software, at least not in its original form. When detected, these AI systems recognize them as dogs, giraffes, or other animals included in the design, not as garments actual people wear on a random Sunday afternoon.
More importantly, the faces of those coming in contact with facial recognition software are not detected. In other words, Cap_able successfully created a textile that can trick AI and help people protect their privacy without covering their facial features or hiding from cameras.
Source: Capable Design
No one has ever been able to draw attention to the misuse of facial recognition software in such a creative way as Cap_able. The company helps people take charge of their privacy and suppress intrusion coming from unwanted parties by creating cozy and stylish clothes, and people are in love with the designs and ideas behind the project. As a result of Cap_able’s effort, the biometric data of their consumers aren’t collected, stored, or misused – of course, only when they’re wearing the garments.
Besides helping its customers, Cap_able’s collection raises awareness of privacy issues, not only in Italy but worldwide as well. As long as the dialogue exists and people discuss ways to avoid violation of privacy and find solutions that would allow us to use facial recognition in a safe and responsible manner, we have to agree that an adequate solution must be on the way.