The use of face verification systems is becoming increasingly common in various aspects of life, from social media to financial transactions. However, people with facial differences are struggling to participate due to the technology’s inability to recognize their unique features.
Autumn Gardiner thought updating her driving license would be straightforward after getting married. She visited the local DMV office in Connecticut and was asked to update her photo. However, every time staff tried to take her photo, the system would reject it. Gardiner believes she is not the only one facing this issue.
Around half a dozen people tell WIRED they are struggling with face verification systems due to their facial differences. These individuals have undergone multiple surgeries and experienced stigma throughout their lives. They claim that face recognition technology is failing them and making it difficult for them to access public services, such as passports, credit scores, and financial institutions.
Phyllida Swift, CEO of Face Equality International (FEI), estimates that there are more than 100 million people worldwide living with facial disfigurements. FEI’s research shows that these individuals face problems with airport passport gates, photo apps, social media video filters, background blurring on video calls, and more.
The issues arise because the underlying machine learning technology is not trained on diverse datasets, which can lead to inaccurate results for people with facial differences. Organizations like Face Equality International are advocating for alternative methods, such as manual verification processes or special accommodations for individuals with disabilities.
As AI-powered face recognition systems become more prevalent, these individuals are facing a lack of inclusivity and accessibility. Many have reported feeling isolated and excluded from society due to online appearance-based bullying.
Some businesses and organizations are starting to acknowledge the issue and provide alternative solutions. However, progress is slow, and more needs to be done to ensure that face verification systems work for everyone.
While there were some positive responses to Gardiner’s experience at the DMV in Connecticut, she has heard stories of people with facial differences facing similar issues in other states. The problem seems insurmountable due to the widespread use of face recognition technology.
In recent updates, Al-Khaled was able to gain access to Social Security Administration services after ID.me provided assistance. However, many more individuals still need help navigating these complex systems.
Source: https://www.wired.com/story/when-face-recognition-doesnt-know-your-face-is-a-face