Super-Recognizer Secrets: Boosting AI Facial Recognition

Published Date : 11/9/2025 

New research from UNSW Sydney reveals how super-recognizers, individuals with exceptional face recognition abilities, can significantly improve AI systems. Their precise gaze patterns provide valuable insights for enhancing facial recognition technologies. 

You might know someone who struggles to recognize people, even if they’re famous and on TV all the time. On the other hand, there are some who seem to always know a face, remembering them from a brief meeting at a party years ago.

Now, new research from UNSW Sydney reveals that the secret to elite face recognition lies in not only taking in someone’s whole face but in what we choose to look at. In a study, researchers uncovered why “super-recognisers” — individuals with exceptional face recognition abilities — outperform the average person.

The answer lies in the precision of their gaze. “Super-recognisers don’t just look harder, they look smarter,” says Dr James Dunn, lead author of the study published in Proceedings of the Royal Society B: Biological Sciences. The study is titled “Super-recognisers sample visual information of superior computational value for facial recognition.”

“They choose the most useful parts of a face to take in,” Dunn continued. “They’re not actually seeing more, instead, their eyes naturally look at the parts of a face that carry the best clues for telling one person from another.”

To decode this visual expertise, researchers used eye-tracking technology to monitor how 37 super-recognisers and 68 average observers scanned facial images. The team then recreated these gaze patterns and fed them into nine pre-trained facial recognition neural networks.

The results were head-turning. When AI systems were guided by the eye movements of super-recognisers, they achieved significantly higher accuracy in matching faces, even when the total visual information was the same. This suggests that the quality, not quantity, of visual input is what boosts recognition performance.

“Even when you control for the fact that they’ve looked at more parts of the face, it turns out what they are looking at is also more valuable for identifying people,” Dunn says. The study says in its abstract that identity matching accuracy improved across all nine deep neural networks (DNNs) when visual information was used as that used by super-recognizers.

The findings have implications for biometric technologies, particularly facial recognition systems used in security, border control, and identity verification. While AI systems like airport eGates scan every pixel of a face under ideal conditions, human recognition still holds an edge in less controlled environments, especially when familiarity and context come into play, the academic explained.

But as AI continues to evolve, that gap is closing. By mimicking the gaze strategies of super-recognisers, future biometric systems could become more efficient and resilient in real-world conditions.

Can the average person train themselves to see like a super-recogniser? Unfortunately, no. According to Dr Dunn, this ability appears to be automatic and deeply embedded in the brain’s visual processing.

“It’s like caricature,” he explains. “Super-recognisers seem to do that visually — they’re tuning in to the features that are most diagnostic about a person’s face.” Dunn says super-recognisers differ from the average person because something goes in their brain related to processing information rather than only just about what they’re looking at in a face.

For those curious about their own facial recognition abilities, UNSW offers a free online test to identify potential super-recognisers. 

Frequently Asked Questions (FAQS): 

Q: What are super-recognizers?

A: Super-recognizers are individuals who have exceptional face recognition abilities, allowing them to remember and recognize faces with high accuracy, even from brief encounters.

Q: How do super-recognizers outperform average people in face recognition?

A: Super-recognizers outperform average people by focusing on the most useful parts of a face, which carry the best clues for identifying individuals, rather than just looking at more parts of the face.

Q: What technology was used in the study to understand super-recognizers?

A: Researchers used eye-tracking technology to monitor how super-recognizers and average observers scanned facial images, and then recreated these gaze patterns to feed into facial recognition neural networks.

Q: What are the implications of this research for facial recognition systems?

A: The findings suggest that by mimicking the gaze strategies of super-recognizers, facial recognition systems can become more efficient and accurate, especially in less controlled environments.

Q: Can anyone train themselves to become a super-recognizer?

A: Unfortunately, no. The ability of super-recognizers appears to be automatic and deeply embedded in the brain's visual processing, making it difficult to train oneself to achieve this level of face recognition. 

More Related Topics :