The science of super-recognizers’ amazing fairies

Editor’s note: This version is a slight modification of a version originally distributed by the University of Wollongong on 1 September.

Super recognizers never forget a face. They can catch a glimpse of their childhood friend in the rearview mirror and instantly know it’s them. They help police departments and security agencies identify suspects. They are also good private detectives and unofficial investigators.

But as fascinating as their superpower is, it remains poorly understood. Until now, scientists believed that super-recognizers were so good with faces because they processed them holistically by taking a snapshot of the face and storing it.

In an article published Aug. 31 in the journal Psychological sciencepsychologists from UNSW Sydney and the University of Wollongong (UOW) challenged this view, proving that super-recognizers, who make up about 2 per cent of society, look at faces just like the rest of us, but do it faster and more accurately.

how does this happen

UNSW researcher and lead study author Dr James Dunn explains that when super-recognizers catch a glimpse of a new face, they break it up into parts and then store them in the brain as composite images.

“They can still recognize faces better than others, even when they can only see smaller regions at a time. This suggests that they can piece together an overall impression from smaller pieces, rather than a holistic impression taken at a single glance,” Dunn said.

For the purpose of the study, co-author Dr Sebastien Miellet, a UOW researcher in the School of Psychology and an expert in active vision, used eye-tracking technology to analyze how the super-recognizers scan and process faces and their parts.

“We can see with great precision not only where people are looking, but which parts of the visual information they are using,” Mielet said.

When studying the visual processing patterns of super-recognizers, Dunn and Mielet realized that, unlike typical recognizers, super-recognizers focused less on the eye region and distributed their gaze more evenly than typical viewers, extracting information from other features of the face, especially when studying faces.

“So the advantage of super-recognizers is their ability to capture highly distinctive visual information and assemble all the parts of the face like a puzzle, quickly and accurately,” Mielet said.

UNSW and UOW researchers will continue to study the super-recognizer population.

Miellet believes one hypothesis is that the superpower of super-recognizers may stem from a specific curiosity and behavioral interest in other people. Potentially, super-recognizers may also be more empathetic than most of us.

“In the next stages of our research, we will equip some super-recognizing and typical viewers with a portable eye-tracking device and put them out on the street to observe, not in the lab, but in real life, how they interact with the world,” Mielet said.

About the research

“Sampling face information in super-recognizers” by Dunn James D*, Varela Victor PL, Nicholls Victoria I, Papinuto Michael, White David, Mielet Sebastien* is published in Psychological science2022. Preprint DOI: 10.31234/

*Equal contribution

About the researcher

Dr Sebastien Miellet is a researcher and lecturer in the UOW School of Psychology and head of the Active Vision Lab, where his research team analyzes oculomotor strategies and how they affect face processing, scene perception, pedestrian safety, reading, etc. n.

Media resources

Dr Sebastien Miellet is available for interviews through the UOW media office.

High resolution images of Dr. Miellet are available for download from Dropbox here:

Media contacts

Alex RezelskaMedia and Public Relations, UOW

M: 0498 964 183 | E: [email protected]

UOW Media Office, T: +61 4221 4227 | Email: [email protected]

To contact Dr. James Dunn, contact Lachlan Gilbert at UNSW Media, T: +61 2 9065 5241

Email: [email protected]

Disclaimer: AAAS and EurekAlert! is not responsible for the accuracy of the news published on EurekAlert! by contributing institutions or for using any information through the EurekAlert system.

Leave a Comment