How does Hyland do it? Nobody knows. But since 2011, about 200 London police officers have been recruited to an elite squad of super recognizers. Officials say they have tripled the number of criminal suspects identified from surveillance photos or on the street each week, and even helped prevent some crimes like muggings, drug deals and assaults.
"When we have an image of an unidentified criminal, I know exactly who to ask instead of sending it out to everyone and getting a bunch of false leads," said Mick Neville, Detective Chief Inspector at Scotland Yard. Neville started the super recognizer unit after realizing the police had no system for identifying criminals based on images, unlike those for DNA and fingerprints.
The unit proved especially valuable after riots hit London in the summer of 2011. After the violence, Scotland Yard combed through hundreds of hours of surveillance video. So far, there have been nearly 5,000 arrests; around 4,000 of those were based on police identifications of suspects from video images. The super recognizers were responsible for nearly 30 percent of the identifications, including one officer who identified almost 300 people. A facial recognition software program made only one successful identification, according to Neville.
Weeks before the Notting Hill Carnival, the biggest street festival in Europe, kicked off last month, the super recognizers were given images of known criminals and gang members. After the carnival began, 17 super recognizers holed up in a control room to study surveillance footage and spot the potential troublemakers.
Once targeted people were identified, police officers were sent to the scene as a pre-emptive strategy. Neville said that likely prevented some crimes like thefts and assaults.
Neville said one super recognizer saw what he thought was a drug deal, but wasn't sure. The next day, the super recognizer saw the same person and when police intervened, they found the suspect with crack cocaine.
He noted that the officers aren't infallible and that their identification is only the start of a case, after which police start looking for other evidence.
Legal authorities warned it could be problematic to use super recognizers as expert witnesses in court, such as in situations where they identify criminals based on an imperfect image.
"Unless we subject them to (rigorous testing), then we are just taking their word on trust and we have no reason to do this," said Mike Redmayne, a law professor at the London School of Economics. "Perhaps they can do what they say, but we don't have the evidence yet," he wrote in an email. "If it was up to me, I would not (allow) it in court."
In the U.S., experts thought it would be up to individual judges to decide whether super recognizers needed to be verified before allowing their testimony in court.
"It's not clear to me that the law will demand they be tested first," said David Kaye, a distinguished professor of law at Penn State. He said the identification skills of super recognizers might be analogous to those of sniffer dogs, whose ability to sniff out drugs are mostly accepted without confirmatory tests. Kaye also noted cases where expert witnesses didn't need to have their skills verified before testifying in court and thought that in most instances, the prosecution would have more evidence than simply the identification of an alleged criminal by a super recognizer.
He said the skills of super recognizers might be more plausibly used in obtaining search warrants.
"There aren't strict rules for getting a warrant," Kaye said. "The judge is supposed to exercise independent judgment but often anything goes," he said, explaining that a super recognizer's identification of a suspect based on a grainy image might be sufficient to issue a search warrant.
Charles Farrier, a spokesman for the U.K. privacy group, No CCTV, called the police's use of super recognizers "the latest gimmick" being used to promote the widespread use of surveillance cameras. According to the group, Britain has the most surveillance cameras per person in the world.