Human plus machine – face recognition at its best

May 29, 2018 by Deborah Smith, University of New South Wales
Human plus machine – face recognition at its best
Do these two faces show the same person? Credit: J. Stoughton/NIST

The first study to compare performances of trained facial examiners, super-recognisers, and facial-recognition algorithms, has revealed a combination of human and computer decision-making is most accurate.

The first study to compare the performances of trained forensic facial examiners, people known as super-recognisers who have a natural talent for face identification, and facial-recognition computer algorithms, has revealed that a combination of human and computer decision-making is most accurate.

The study, by a team of scientists from the National Institute of Standards and Technology in the US and three universities including UNSW Sydney, is published in the Proceedings of the National Academy of Sciences.

"Experts in face identification often play a crucial role in criminal cases," says study team member, UNSW psychologist Dr. David White.

"Deciding whether two images are of the same person, or two different people, can have profound consequences.

"When facial comparison evidence is presented in court, it can determine the outcome of a criminal trial. Errors on these decisions can potentially set a guilty person free, or wrongly convict an innocent person," he says.

The international study involved a total of 184 participants from five continents—a large number for an experiment of this type.

Eighty-seven were trained professional facial examiners, while 13 were super-recognisers—people with exceptional natural ability, but no training. The remaining 84 were control participants with no special training or natural ability, including 53 fingerprint examiners and 31 undergraduate students.

Participants received pairs of face images and rated the likelihood of each pair being the same person on a seven-point scale. The research team intentionally selected extremely challenging pairs, using images taken with limited control of illumination, expression and appearance.

They then tested four of the latest computerised facial recognition algorithms, all developed between 2015 and 2017, using the same image pairs.

"As a group, trained forensic examiners outperformed the other groups," says Dr. White.

"Another important insight from the study was that the most advanced facial-recognition algorithms are now as accurate as the very best humans.

"However, the results with people showed large variation in accuracy of individuals in all the groups tested. This ranged from near random guessing, with an accuracy of about 50%, to a perfect score of 100%.

"This variability is a problem, because it is common practice for just one examiner to present decisions in court," says Dr. White.

The study found that combining several examiners' opinions produced higher accuracy than one examiner working alone, and led to less variability in accuracy compared to individual responses.

"But the surprising best solution to the problem of individual variability is to combine the responses from one examiner with the responses from the best . A combination of human and computer decision-making leads to the most accurate results," Dr. White says.

Explore further: Study shows face recognition experts perform better with AI as partner

More information: P. Jonathon Phillips el al., "Face recognition accuracy of forensic examiners, superrecognizers, and face recognition algorithms," PNAS (2018). www.pnas.org/cgi/doi/10.1073/pnas.1721355115

Related Stories

Study shows face recognition experts perform better with AI as partner

May 28, 2018
Experts at recognizing faces often play a crucial role in criminal cases. A photo from a security camera can mean prison or freedom for a defendant—and testimony from highly trained forensic face examiners informs the jury ...

Forensic examiners pass the face matching test

September 1, 2015
The first study to test the skills of FBI agents and other law enforcers who have been trained in facial recognition has provided a reassuring result - they perform better than the average person or even computers on this ...

Multi-racial facial recognition system provides more accurate results, study says

November 6, 2017
A multi-racial facial recognition system delivers more accurate results than those typically used today, a new study published in Pattern Recognition journal has revealed.

Facial expressions can cause us problems in telling unfamiliar faces apart

June 2, 2017
Using hundreds of faces of actors from movies, psychologists from the University of Bristol have shown how facial expressions can get in the way of our ability to tell unfamiliar faces apart.

Recommended for you

FDA approves brain stimulation device for OCD

August 17, 2018
(HealthDay)—A brain stimulation device to treat obsessive-compulsive disorder (OCD) has received approval for marketing Friday by the U.S. Food and Drug Administration.

People are more honest when using a foreign tongue, research finds

August 17, 2018
New UChicago-led research suggests that someone who speaks in a foreign language is probably more credible than the average native speaker.

16 going on 66: Will you be the same person 50 years from now?

August 17, 2018
How much do you change between high school and retirement? The answer depends on whether you're comparing yourself to others or to your younger self.

Research eyes role of stress in mental illnesses

August 17, 2018
We all face stress in our lives. Even researchers seeking to understand why some people shrug it off while others face battles against disorders like depression or PTSD.

It's okay when you're not okay: Study re-evaluates resilience in adults

August 16, 2018
Adversity is part of life: Loved ones die. Soldiers deploy to war. Patients receive terminal diagnoses.

Expecting to learn: Language acquisition in toddlers improved by predictable situations

August 16, 2018
The first few years of a child's life are crucial for learning language, and though scientists know the "when," the "how" is still up for debate. The sheer number of words a child hears is important; that number predicts ...

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.