GeoDA: a geometric framework for black-box adversarial attacks


METADATA ONLY
Loading...

Date

2020

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric
METADATA ONLY

Data

Rights / License

Abstract

Adversarial examples are known as carefully perturbed images fooling image classifiers. We propose a geometric framework to generate adversarial examples in one of the most challenging black-box settings where the adversary can only generate a small number of queries, each of them returning the top-1 label of the classifier. Our framework is based on the observation that the decision boundary of deep networks usually has a small mean curvature in the vicinity of data samples. We propose an effective iterative algorithm to generate query-efficient black-box perturbations with small p norms which is confirmed via experimental evaluations on state-of-the-art natural image classifiers. Moreover, for p=2, we theoretically show that our algorithm actually converges to the minimal perturbation when the curvature of the decision boundary is bounded. We also obtain the optimal distribution of the queries over the iterations of the algorithm. Finally, experimental results confirm that our principled black-box attack algorithm performs better than state-of-the-art algorithms as it generates smaller perturbations with a reduced number of queries.

Publication status

published

Editor

Book title

2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Journal / series

Volume

Pages / Article No.

8443 - 8452

Publisher

IEEE

Event

2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2020) (virtual)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Organisational unit

09462 - Hofmann, Thomas / Hofmann, Thomas check_circle

Notes

Due to the Coronavirus (COVID-19) the conference was conducted virtually.

Funding

Related publications and datasets