Chicago Robber Caught By Facial Recognition Sentenced To 22 Years

mpicpp (3454017) writes with this excerpt from Ars: "The first man to be arrested in Chicago based on facial recognition analysis was sentenced last week to 22 years in prison for armed robbery. ... In February 2013, Pierre Martin robbed a man at gunpoint while on a Chicago Transit Authority (CTA) train. After taking the man's phone, Martin jumped off the train. However, his image was captured by CTA surveillance cameras and was then compared to the Chicago Police Department's database of 4.5 million criminal booking images. Martin, who already had priors, had a mugshot in the database. He was later positively identified by witnesses. At trial, Martin also admitted to committing a similar robbery also on the Pink Line in January 2013—his face was captured during both robberies."

Comments

Watch_Dogs

[Travis Mansbridge]

Shoulda just hacked the Chicago camera system with his phone.

Fingerprints

[Dan East]

This is nothing more than the type of fingerprint matching that's been going on for many decades. It just puts a name to a person after the fact. Now on the other hand, if he was actively recognized via facial recognition as he was out and about in public and then apprehended, well that would be a different story.

Re:Fingerprints

[Anubis IV]

I agree, but I think there's another concern here as well: false positives are significantly more dangerous than with other fingerprinting techniques. If DNA samples or fingerprints provide false positives, we have (admittedly error-prone) eyewitnesses as a final layer of defense, and since people who look entirely different can have similar fingerprints or DNA signatures, it's likely that the people look nothing alike. Not so with facial recognition, since a false positive is likely to be close enough to a true positive that it will be incorrectly affirmed by eyewitnesses, even if the authorities don't bias them by telling them that the guy was a match.

None of which is to say that I think we should stop using it, since it is a valuable tool. I merely think that it needs to be used with an understanding of its faults and taken with the grain of salt it deserves.

FTFY

[jklovanc]

Chicago Robber Identified By Facial Recognition Sentenced To 22 Years

Caught would imply that he was walking down the street and facial recognition directed authorities to him. That did not happen.

Re:My two cents

[TWX]

Any remotely competent lawyer would get that kind of identification thrown out of court. Any lineup, even a photo lineup, without multiple options is inadmissible in court.

Sounds to me like this was used as an investigative lead that helped them find other evidence, rather than as the principal evidence presented in court. This really isn't different than a police officer viewing the recording to see the offender's face, then going through books of mugshots to find the face, then investigating those people that the officer thinks might be the offender. This is simply the computer taking the image that the police officer identified and searching those "books" for close matches, then the police looking at the MO of the crime as compared to the MO of the person previously arrested, and investigating ones that have the most commonality first.

In this case they identified a suspect, the suspect apparently had offended in this same way before, and the suspect was tried and convicted. This doesn't seem to violate any new privacy considerations- the recordings being collected themselves are nothing new, and the mugshot database isn't either. Simply making the comparison itself doesn't add any new fuel to the fire of personal liberty complaints or of violation of privacy.