Idemia says NYPD bomb scare suspect ID shows facial recognition best practices

Facial recognition enabled New York Police to quickly identify a man who allegedly placed rice cookers initially suspected of being bombs in the city’s subway system, in a process Idemia says demonstrated best practices for ensuring biometric accuracy and mitigating bias.

The New York Post reports that police were alerted of suspicious devices being left in Fulton Street subway station at around 7:15 a.m. on August 16, and detectives from the Facial Identification Section had distributed digital images of the suspect to every officer in the department by 8:15 a.m. It would not be until after 9:00 a.m. that the NYPD bomb squad determined the rice cookers had not been modified to explode.

Idemia provides the NYPD’s facial biometric technology, along with DataWorks Plus, according to the Post, and in a LinkedIn post Idemia outlines four key steps taken during the process. The NYPD collected multiple images from subway security cameras, and reviewed “a few hundred potential matches” after searching its mugshot database. Once a potential suspect had been identified, they collected additional images on social media to check for any corroborating or disqualifying features, and then had a second trained Face Analyst confirm the match.

“Five years, ago you probably have endless detectives looking through videos … images of arrested individuals based on descriptions,” Sgt. Edwin Coello, head of the NYPD facial recognition unit told the Post. “It could take several hours or several days. This is the most important type of case that we’d see out there: a possible terrorist attack in NYC.”

The police department has received criticism for not publishing a comprehensive policy on the technology’s use, and has battled Georgetown Center on Privacy and Technology over providing documentation related to the system. New York City Council is also considering a bill to provide transparency around the use of surveillance technology, which the NYPD opposes.

“To not use technology like this would be negligent. This is the future,” said Chief of Detectives Dermot Shea. “We are putting in place a system where there’s checks and balances to make sure we do our due diligence to make sure it’s a high probability the person who we are looking for.”

Article Topics

accuracy  |  best practices  |  biometrics  |  criminal ID  |  DataWorks Plus  |  IDEMIA  |  matching  |  police