September 27, 2017 Robust Physical-World Attacks on Machine Learning Models
Speaker: Earlence Fernandes
Abstract: Deep neural network-based classifiers are known to be vulnerable to adversarial examples that can fool them into misclassifying their input through the addition of small-magnitude perturbations. However, recent studies have demonstrated that such adversarial examples are not very effective in the physical world—they either completely fail to cause misclassification or only work in restricted cases where a relatively complex image is perturbed and printed on paper. In this paper we propose a new attack algorithm—Robust Physical Perturbations (RP2)— that generates perturbations by taking images under different conditions into account. Our algorithm can create spatiallyconstrained perturbations that mimic vandalism or art to reduce the likelihood of detection by a casual observer. We show that adversarial examples generated by RP2 achieve high success rates under various conditions for real road sign recognition by using an evaluation methodology that captures physical world conditions. We physically realized and evaluated two attacks, one that causes a Stop sign to be misclassified as a Speed Limit sign in 100% of the testing conditions, and one that causes a Right Turn sign to be misclassified as either a Stop or Added Lane sign in 100% of the testing conditions.
Note: Since this is our first meeting of the quarter, we will have a three-minute organizational meeting before the talk begins.
October 04, 2017 TBA
October 11, 2017 Matt Kraning (Quadium)
Speaker: Matt Kraning (Quadium)
Abstract:
October 18, 2017 TBA
Speaker:
Abstract:
October 25, 2017 TBA
Speaker:
Abstract:
November 01, 2017 TBA
November 08, 2017 TBA
Speaker:
Abstract:
November 15, 2017 TBA
Speaker:
Abstract:
November 23, 2017 Thanksgiving Week
November 29, 2017 TBA
Speaker:
Abstract:
December 06, 2017 TBA
Speaker:
Abstract:
December 13, 2017 TBA
Speaker:
Abstract: