Copy and Paste: A Simple But Effective Initialization Method for Black-Box Adversarial Attacks

06/14/2019
by   Thomas Brunner, et al.
0

Many optimization methods for generating black-box adversarial examples have been proposed, but the aspect of initializing said optimizers has not been considered in much detail. We show that the choice of starting points is indeed crucial, and that the performance of state-of-the-art attacks depends on it. First, we discuss desirable properties of starting points for attacking image classifiers, and how they can be chosen to increase query efficiency. Notably, we find that simply copying small patches from other images is a valid strategy. In an evaluation on ImageNet, we show that this initialization reduces the number of queries required for a state-of-the-art Boundary Attack by 81 black-box adversarial examples.

READ FULL TEXT

page 1

page 3

page 4

research
08/19/2019

Hybrid Batch Attacks: Finding Black-box Adversarial Examples with Limited Queries

In a black-box setting, the adversary only has API access to the target ...
research
03/13/2020

GeoDA: a geometric framework for black-box adversarial attacks

Adversarial examples are known as carefully perturbed images fooling ima...
research
07/13/2020

Hard Label Black-box Adversarial Attacks in Low Query Budget Regimes

We focus on the problem of black-box adversarial attacks, where the aim ...
research
02/19/2019

There are No Bit Parts for Sign Bits in Black-Box Attacks

Machine learning models are vulnerable to adversarial examples. In this ...
research
01/22/2021

Generating Black-Box Adversarial Examples in Sparse Domain

Applications of machine learning (ML) models and convolutional neural ne...
research
01/21/2019

Perception-in-the-Loop Adversarial Examples

We present a scalable, black box, perception-in-the-loop technique to fi...
research
02/21/2021

A Zeroth-Order Block Coordinate Descent Algorithm for Huge-Scale Black-Box Optimization

We consider the zeroth-order optimization problem in the huge-scale sett...

Please sign up or login with your details

Forgot password? Click here to reset