Presented By O’Reilly and Intel AI
Put AI to Work
April 29-30, 2018: Training
April 30-May 2, 2018: Tutorials & Conference
New York, NY

Fooling neural networks in the physical world

Andrew Ilyas (Massachusetts Institute of Technology), Logan Engstrom (Massachusetts Institute of Technology), Anish Athalye (Massachusetts Institute of Technology)
11:55am–12:35pm Wednesday, May 2, 2018
Interacting with AI, Models and Methods
Location: Concourse A
Average rating: *****
(5.00, 2 ratings)

Who is this presentation for?

  • Those concerned about security in machine learning

Prerequisite knowledge

  • A basic understanding of neural networks, image classification, and backpropagation

What you'll learn

  • Explore an approach that produces adversarial examples that fool neural networks at any angle in the physical world


Adversarial examples generated with standard methods do not consistently fool a classifier in the physical world, due to a combination of viewpoint shifts, camera noise, and other natural transformations. These examples require complete control over direct input to the classifier, which is fundamentally impossible in many real-world systems.

Andrew Ilyas, Logan Engstrom, and Anish Athalye offer an overview of an algorithm that produces adversarial examples that remain adversarial under an attacker-chosen distribution and demonstrate its application in two dimensions, producing adversarial images that are robust to noise, distortion, and affine transformation and showing that these input distortions are ineffective against robust adversarial examples. They then apply the algorithm to produce the first physical 3D-printed adversarial objects, demonstrating how the approach works in the real world.

Photo of Andrew Ilyas

Andrew Ilyas

Massachusetts Institute of Technology

Andrew Ilyas is an undergraduate student at the Massachusetts Institute of Technology.

Photo of Logan Engstrom

Logan Engstrom

Massachusetts Institute of Technology

Logan Engstrom is an undergraduate student at the Massachusetts Institute of Technology.

Photo of Anish Athalye

Anish Athalye

Massachusetts Institute of Technology

Anish Athalye is a graduate student at the Massachusetts Institute of Technology.