Subscribe now

Technology

Image-generating AI creates uncanny optical illusions

Generative AI models designed to create images from text can be adapted to construct complex optical illusions - pictures that look like different things when viewed from far away or with motion blur

By Matthew Sparkes

29 April 2024

Motion hybrid - ancient ruins and a teddybear

An AI-generated image of ancient ruins looks like a teddy bear when motion blur is added

Daniel Geng, Inbum Park, Andrew Owens

AI models designed to create images from text descriptions can make optical illusions that look like different objects depending on how they are viewed.

Daniel Geng at the University of Michigan and his colleagues made slight adaptions to an existing AI model to get it to make a range of illusions. Some appear to depict other things as you view them from further away; some become three different images depending on their rotation and…

Sign up to our weekly newsletter

Receive a weekly dose of discovery in your inbox! We'll also keep you up to date with New Scientist events and special offers.

Sign up

To continue reading, subscribe today with our introductory offers

View introductory offers

No commitment, cancel anytime*

Offer ends 2nd of July 2024.

*Cancel anytime within 14 days of payment to receive a refund on unserved issues.

Inclusive of applicable taxes (VAT)

or

Existing subscribers

Sign in to your account