Short Training Helps People Spot AI Faces in the Battle Against Deepfake Fraud
Researchers found that training on six facial markers nearly doubled participants’ accuracy in spotting AI-generated faces.
- On Tuesday, Australian National University researchers published a study in PNAS showing brief training sessions significantly improve human ability to detect AI-generated faces.
- As AI-generated faces become more realistic, traditional signs of fakery have vanished. Associate Professor Amy Dawel noted these images are often 'more symmetrical, proportional and attractive,' making detection difficult.
- Training focuses on six markers: distinctiveness, memorability, proportionality, symmetry, attractiveness and expressiveness. Participants' accuracy in detecting AI faces increased from 40 per cent to 80 per cent.
- Researchers at the University of Victoria, led by psychology professor Jim Tanaka and post-doctoral fellow Eric Mah, replicated these findings, confirming detection capabilities can be effectively trained.
- Future research will verify if these detection skills generalize to other AI-generated faces beyond StyleGAN, as Dawel emphasized such skills are essential for maintaining online safety.
10 Articles
10 Articles
The Free Press
UVic researchers looking to help people spot deepfake AI images
A recent University of Victoria study suggests that humans have been successfully trained to spot computer-generated faces from from photos of real human faces – also known as deepfakes. The UVic Different Minds Lab team, led by psychology professor Jim Tanaka and post-doctoral fellow Eric Mah, partnered with the Australian National University (ANU) to develop a quick, effective and robust training technique for improving people’s detection of d…
As Deepfake Scams Surge, Researchers Test Whether People Can be ‘Trained’ to Spot Fake Faces
The rapid rise of deepfakes is making it increasingly difficult for people to tell what is real online, but new research from the Australian National University suggests humans can be trained to better detect AI-generated faces at a time when digital impersonation and fraud are escalating. Researchers from the ANU Emotions and Faces Lab found that targeted training significantly improved people’s ability to distinguish between real and AI-genera…
Short training helps people spot AI faces in the battle against deepfake fraud
Humans have been successfully trained to spot AI-generated faces in a study led by researchers at the Australian National University (ANU) Emotions and Faces Lab. The study, "Training Humans to Detect AI-generated Faces," is published in PNAS.
Coverage Details
Bias Distribution
- 78% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
