adversarial images
Hackers Fool Facial Recognition Into Thinking I’m Mark Zuckerberg
Using a new technique, researchers say they can make AI systems misidentify people by adding small bits of data to the images.
Researchers Tricked AI Into Doing Free Computations It Wasn't Trained to Do
Google researchers demonstrated that a neural network could be tricked into performing free computations for an attacker. They worry that this could one day be used to turn our smartphones into botnets by exposing them to images.
How Cellphone Camera Images Can Fool Machine Vision
Grainy images that would be clear enough to human eyes confuse a machine.
Advertisement