Face recognition ability
Why face recognition ability varies — and how it can be measured in real-world contexts.
Research
How identity decisions are made — and where they go wrong
We study how people recognise faces, remember events, and judge identity — especially in high-stakes settings like policing, security, and digital systems where mistakes have real consequences.
Used by 500,000+ participants · Applied in policing and intelligence · Award-winning research
Overview
Decisions about identity are difficult — and often high-stakes.
Every day, people make decisions about identity — comparing faces, recalling events, deciding whether two images show the same person, or judging whether a face is real or AI-generated.
These decisions are often made under uncertainty, yet they carry real consequences: wrongful identification, missed matches, or misplaced trust in AI systems.
Our research asks a simple question: when are people accurate, when are they not, and how can these decisions be improved?
Research themes
Core areas of investigation
Memory and confidence
How stress and confidence shape memory reports, evidence, and identity decisions.
AI-generated faces
Why synthetic identities are difficult to detect, and what this means for online trust.
Human–AI interaction
How AI changes identity decisions — and how systems can be designed to reduce error.
Research innovations
Combining cognitive science with emerging methods
Eye-tracking and behaviour
Tracking where people look when comparing faces to understand decision strategies.
Machine learning and AI
Modelling decision processes and studying how people interact with AI-generated identities.
Large-scale behavioural testing
Measuring variation in cognitive ability and decision-making across large participant samples.
Key findings
What the research shows
Ability varies dramatically
Some people are exceptional with faces, while others struggle with the same tasks.
People are often overconfident
People tend to overestimate their ability to detect AI-generated identities.
AI changes decision-making
AI systems can change how people interpret faces and judge identity.
Expert selection improves outcomes
Identifying high-performing people can improve accuracy in applied settings.
Impact
From cognitive science to real-world practice
Research from the lab is used in policing, intelligence, and public-facing tools, with measurable impact across real-world systems.
Policing and security
Research has informed personnel selection and deployment for specialist face identification roles.
Policy and international impact
Briefings and collaborations support identity verification policy and applied AI discussions.
Evidence and memory in practice
Research on memory under stress informs structured recall protocols for policing contexts.
Approach
Studying decisions in realistic conditions
We design tasks that reflect real-world decisions: comparing faces, recalling events, and identifying people under uncertainty.
This includes controlled experiments, large-scale online testing, and collaborations with organisations that make identity decisions in practice.
Try the research tasks:
By combining laboratory precision with applied settings, we identify where errors occur — and how they can be reduced.
Next step
Interested in working together?
We collaborate with researchers, students, industry, and government on applied problems in identity, decision-making, and AI.