Adversarial Examples Are Not Easily Detected
Why is this work in the frame?
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
No Canadian affiliation. An affiliation-only frame — the usual design — would never have seen this work. It is one of the works that make the case for inverting the frame.
Abstract
Neural networks are known to be vulnerable to adversarial examples: inputs that are close to natural inputs but classified incorrectly. In order to better understand the space of adversarial examples, we survey ten recent proposals that are designed for detection and compare their efficacy. We show that all can be defeated by constructing new loss functions. We conclude that adversarial examples are significantly harder to detect than previously appreciated, and the properties believed to be intrinsic to adversarial examples are in fact not. Finally, we propose several simple guidelines for evaluating future proposed defenses.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
The record
- Venue
- Topic
- Adversarial Robustness in Machine Learning
- Field
- Computer Science
- Canadian institutions
- —
- Funders
- Air Force Office of Scientific ResearchMultidisciplinary University Research InitiativeCanadian Institute for Advanced ResearchWilliam and Flora Hewlett Foundation
- Keywords
- Adversarial systemComputer scienceArtificial intelligenceSimple (philosophy)Deep neural networksSpace (punctuation)Artificial neural networkMachine learningTheoretical computer scienceEpistemology
- Has abstract in OpenAlex
- yes