We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Focused Adversarial Attacks

Abstract: Recent advances in machine learning show that neural models are vulnerable to minimally perturbed inputs, or adversarial examples. Adversarial algorithms are optimization problems that minimize the accuracy of ML models by perturbing inputs, often using a model's loss function to craft such perturbations. State-of-the-art object detection models are characterized by very large output manifolds due to the number of possible locations and sizes of objects in an image. This leads to their outputs being sparse and optimization problems that use them incur a lot of unnecessary computation.
We propose to use a very limited subset of a model's learned manifold to compute adversarial examples. Our \textit{Focused Adversarial Attacks} (FA) algorithm identifies a small subset of sensitive regions to perform gradient-based adversarial attacks. FA is significantly faster than other gradient-based attacks when a model's manifold is sparsely activated. Also, its perturbations are more efficient than other methods under the same perturbation constraints. We evaluate FA on the COCO 2017 and Pascal VOC 2007 detection datasets.
Subjects: Machine Learning (cs.LG); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Neural and Evolutionary Computing (cs.NE)
Cite as: arXiv:2205.09624 [cs.LG]
  (or arXiv:2205.09624v1 [cs.LG] for this version)

Submission history

From: Charles Fleming [view email]
[v1] Thu, 19 May 2022 15:38:23 GMT (2427kb,D)

Link back to: arXiv, form interface, contact.