Detroit police change facial recognition procedure as falsely accused woman sues

Police leadership in the U.S. city of Detroit is reacting to a wrongful arrest lawsuit by defending facial recognition software and coming down hard on a detective. The department is also changing its procedural policies for facial recognition use.

In February, an eight-months pregnant woman was arrested and detained in a cell for 11 hours, accused of robbery and carjacking. That woman, Porcha Woodruff, was cleared a month later. She is suing the Detroit Police Department for wrongful arrest and imprisonment.

From the department’s point of view, the problem rests with how police handled a subsequent photo lineup, and policies are being updated.

Woodruff says she was falsely identified by a face biometrics program used by the Detroit police since 2017.

Police Chief James White yesterday held a press conference in which he said that facial recognition was involved in apprehending the plaintiff, but that it did what it was designed to do.

The lead detective in the Woodruff case, who White repeatedly cited in the meeting, knew that witnesses did not report that the carjacker was pregnant. According to the department, the detective did not go back to witnesses about this discrepancy.

Making matters worse, the carjacking victim was shown a photo lineup that included the reference image of Woodruff that the software matched with a photo from a surveillance camera at the scene of the crime.

According to the Detroit Free Press, White said that violates department policies. It is like putting “a suspect’s twin in a lineup.” Doing that can influence a witness’ judgment.

That prohibition stands, but now, two captains have to review request for warrants when facial recognition algorithms are used in the investigation. Also, cops have to use a sequential double-blind lineup. And police involved in an investigation cannot do the lineup procedure.

Detroit has been one of the centers of concern in the United States about the police use of facial recognition.

Article Topics

accuracy  |  best practices  |  biometric-bias  |  biometrics  |  Detroit  |  facial recognition  |  false arrest  |  police  |  United States