After a surge in face mask selfies caused by the COVID-19 pandemic, the AI industry is fighting to keep up and improve their recognition algorithms.
Just recently, CNET discovered thousands of selfies in public data sets presumably to help train AI. In April, researchers published a COVID19 Mask Image Dataset to Github, which holds over 1,200 images from Instagram. The creators behind it used their startup, Workaround, to sift through the database and label photos as “with masks” or “without masks.”
Of course, using online photos to train their programs isn’t new to facial recognition companies. As more and more states start to mandate the use of face masks, the decline in the spread of facial recognition is evident. The existing technology struggles to recognize features when the garment blocks key parts of the face.
“The greatest amount of biometric data that uniquely sets us apart resides in the central portion of the face, just above the brow line all the way down to the chin,” said Eric Hess, senior director of product management for face recognition at facial recognition company SAFR. “When we put on face masks, we are blocking access to a significant amount of data points that help us differentiate one person from another.”
However, developer use of online photos to train AI because of the pressure of the “new normal”, i.e. wearing face masks almost every minute of being outside, is met with criticisms on public safety and ethics.
“People might not like the idea that their picture could be used to develop a database that could go to law enforcement or government surveillance in a foreign autocratic country like China,” said Jake Laperruque, a senior counsel at the Constitution Project. “You’re putting photos out there, maybe not with an expectation of privacy, but you have an expectation of how it can and can’t be used.”