TechTheDay

In a story that could encourage the use of technology or raise doubts, facial recognition tech recently made it to the news being instrumental to the conviction of a thief. Neoface, a facial recognition system purchased by the Chicago Transit Authority for $5.4 million, has sent to jail its first ever convict.

Pierre D. Martin is the first person identified by a facial recognition system used by the police force (Note: person in the photo above is not the thief). As reported on the Chicago Sun Times, the armed robber was sentenced to 22 years of incarceration with facial recognition tech playing a prominent and fundamental role. He was given the verdict in a bench trial in May by Judge Maura Slattery. He was deemed guilty of a robbery he committed in February 2013.

The judge found enough evidence to believe that Martin was guilty of robbing a person on board a Pink Line train. As narrated by his 20-year-old victim, Martin approached him from the back, drew his gun, and ordered him to give his mobile phone. Martin then jumped out of the train. Martin was also involved in stealing at gunpoint a mobile phone from another person in January 2013 at the same Pink Line train.

Fortunately, the train station had surveillance cameras through which Martin’s face was captured in both of the robberies. Because of the multi-million dollar facial recognition technology, Martin was quickly identified. His image (as taken by the surveillance cameras) was compared to millions of other criminals’ faces and he ranked as the closest match or most likely suspect. Of course, the victims and some witnesses verified his identity later on.

It became even easier to identify Martin because his face was already in the police records, with his long record of arrests. In 2006, he was put on probation for owning a stolen motor vehicle. In 2009, he was sent to jail to serve a three-year term for committing an aggravated unlawful use of a weapon. These are just some of the crimes committed by Martin.

Is the Technology Really Reliable?

Is it a good idea to rely on facial recognition tech in solving crimes? As demonstrated in Martin’s case, there appears to be nothing wrong in depending on the technology to facilitate identification and subsequent arrests. If the concern is the accuracy, it’s important to emphasize that law enforcers don’t just arrest people because they are found to be a close match on the facial recognition software. Verifications are still being undertaken and inputs from real people are still included in the identification process.

Facial recognition works most of the time as long as the quality of the images available are clear and detailed enough. Even Facebook’s facial recognition system is already touted to be as accurate as the human brain. The technology has matured well enough that it can facilitate the identification efforts of law enforcers. Again, it’s worth emphasizing that facial recognition won’t be the sole basis for identifying and convicting criminals so if it ends up to be not so accurate, there are safeguards in place to prevent erroneous identification.

Privacy Issues

Unfortunately, with this kind of technology becoming acceptable to the public because of its supposed crime-busting benefits, it’s easy for privacy to be swept under the rug. The system can be used to collect images of ordinary citizens without their knowledge. It can lead to identification without permission and justification. It’s understandable for police mug shots to be used as references but once the government decides to tap on photos shared on social media, a major privacy issue emerges. The benefits are clear but the protection of individual privacy rights may have to be relegated to a lower priority as permission-less facial recognition may become acceptable in the guise of law enforcement.

It is safe to say that there’s nothing wrong with relying on facial recognition tech in finding and prosecuting criminals. After all, it’s not used as a deciding factor but more of a tool for narrowing down search parameters. It is very unlikely for facial recognition to pave the way for unjustifiable or mistaken arrests since it’s very easy to do verifications with the help of complainants and other witnesses unless there is really a concerted effort to pin someone down with false accusations and connivance.