Ability varies widely
Some people are unusually accurate with faces, while others find the same tasks difficult.
Tools & Demos
How good are you at recognising faces?
Our public tools let you experience the kinds of tasks used in research on face recognition, identity decisions, and AI-generated faces.
Used by 500,000+ participants ยท Scientifically validated tasks
Try the tools
Interactive tasks
Test your own performance using the same tasks used in research.
UNSW Face Test
How good are you with faces?
Measure your ability to recognise unfamiliar faces and compare your performance with others.
The top 2% of people are considered super-recognisers
Start test
AI Face Demo
Can you detect an AI-generated face?
Test whether you can distinguish real faces from synthetic identities.
Most people perform worse than they expect.
Try demoAbout the tools
Research tasks, made public
Face recognition and AI detection can feel intuitive, but performance varies widely across people and tasks.
These tools make the science accessible by letting users test their own performance and compare confidence with accuracy.
The results help support research into individual differences, decision-making, and the real-world limits of identity judgements.
What these tools show
Tests reveal meaningful differences
Confidence is often unreliable
People can feel certain about decisions even when accuracy is lower than expected.
AI faces are harder than expected
Synthetic faces are increasingly realistic, making detection harder for many people.
Testing reveals real performance
Structured tasks show patterns that intuition alone cannot.
About the tools
Research tasks, made public
Face recognition and AI detection can feel intuitive, but performance varies widely across people and tasks.
These tools make the science accessible by letting users test their own performance and compare confidence with accuracy.
The results help support research into individual differences, decision-making, and the real-world limits of identity judgements.
Why it matters
Better measurement leads to better decisions
Mistakes in face recognition and identity decisions can have serious consequences in policing, security, digital systems, and everyday online interactions.
By measuring performance directly, these tools help people understand the limits of perception, memory, and confidence.
They also support research translation by connecting public participation with applied questions about identity, evidence, and AI.
Next step
Learn how the tools connect to the research
Explore how face recognition tests and AI face demos support research on identity, evidence, and decision-making.