Scientists employed more than 1,700 pics of students and others without their consent for a facial recognition research backed by US intelligence and military services, as per various media reports. While technically lawful, it has lifted doubts about privacy surrounding facial recognition technology, particularly taking onto consideration how the pics may end up being employed. “This is necessarily normalizing peeping Tom culture,” David Maas of Electronic Frontier Foundation claimed to the media in an interview.
The research, conducted in 2013 and 2012, was made to decide, partly, if algorithms can verify facial features from a remote distance, in poor light and through obstacles. A telephoto camera was installed at a distance of almost 150 Meters with a lot of foot traffic away from a public area. Most of the subjects were seeing away from the camera, with many seeing having their eyes on phones.
After scientists carefully mixed the pics, they made a dataset dubbed as “Unconstrained College Students.” Since it is less detailed and more random than other databases, it can assist researchers design algorithms that work from long distance. That can be specifically employed to the military, for example, by assisting them to see if an incoming boat was a foe or friend.
On a similar note, a teen from New York has lodged a $1 Billion court case against Apple over a fake arrest he claims took place due to what he thinks to be face recognition system by Apple. NYPD cops took Ousmane Bah in custody on November 29, 2018, after he was incorrectly connected to a series of thefts in Delaware, New Jersey, and Manhattan in Apple Store. It seems that, the actual perpetrator employed a stolen ID that had his address, name, and other personal data. On the other hand, since the ID did not have a pic, the court case claims Apple programmed face recognition system of its stores to link the actual thief’s face with Bah’s info.