Data to 7 Oct 2026

The Odds are Odd: A Statistical Test for Detecting Adversarial Examples

arXiv 1902.04818 ↗First posted 13 Feb 2019cs.LG stat.MLcited by 75

Not on the map. Why?

Abstract (arXiv)

We investigate conditions under which test statistics exist that can reliably detect examples, which have been adversarially manipulated in a white-box attack. These statistics can be easily computed and calibrated by randomly corrupting inputs. They exploit certain anomalies that adversarial attacks introduce, in particular if they follow the paradigm of choosing perturbations optimally under p-norm constraints. Access to the log-odds is the only requirement to defend models. We justify our approach empirically, but also provide conditions under which detectability via the suggested test statistics is guaranteed to be effective. In our experiments, we show that it is even possible to correct test time predictions for adversarial attacks with high accuracy.

Key facts. Cited by 75 papers in the map (data through 7 Oct 2026). First posted on arXiv 13 Feb 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 33; 2022: 18; 2023: 9; 2024: 10; 2025: 3; 2026: 2.

75 papers.

Sort:

By area:

2021Advances in adversarial attacks and defenses in computer vision: A surveycited by 127
2021Adversarial Attacks are Reversible with Natural Supervisioncited by 47
2021Detecting Adversarial Examples Is (Nearly) As Hard As Classifying Themcited by 46
2021Neural Mean Discrepancy for Efficient Out-of-Distribution Detectioncited by 42
2023Did You Train on My Dataset? Towards Public Dataset Protection with Clean-Label Backdoor Watermarkingcited by 37
2021Evading Adversarial Example Detection Defenses with Orthogonal Projected Gradient Descentcited by 34
2021A Review of Adversarial Attack and Defense for Classification Methodscited by 29
2022iDECODe: In-distribution Equivariance for Conformal Out-of-distribution Detectioncited by 27
2022"That Is a Suspicious Reaction!": Interpreting Logits Variation to Detect NLP Adversarial Attackscited by 24
2021Indicators of Attack Failure: Debugging and Improving Optimization of Adversarial Examplescited by 23
2022On the Limitations of Stochastic Pre-processing Defensescited by 21
2021Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamicscited by 19

38 more

2024MirrorCheck: Efficient Adversarial Defense for Vision-Language Modelscited by 19
2021Attack as Defense: Characterizing Adversarial Examples using Robustnesscited by 14
2021Two Coupled Rejection Metrics Can Tell Adversarial Examples Apartcited by 14
2023A LLM Assisted Exploitation of AI-Guardiancited by 14
2022Increasing Confidence in Adversarial Robustness Evaluationscited by 12
2021Sample Efficient Detection and Classification of Adversarial Attacks via Self-Supervised Embeddingscited by 11
2022Out-of-Distribution Detection with Reconstruction Error and Typicality-based Penaltycited by 11
2021Resilient Machine Learning for Networked Cyber Physical Systems: A Survey for Machine Learning Security to Securing Machine Learning for CPScited by 10
2021Interpolated Joint Space Adversarial Training for Robust and Generalizable Defensescited by 10
2021SoK: Machine Learning Governancecited by 9
2022Generalizing Adversarial Explanations with Grad-CAMcited by 9
2021NoiLIn: Improving Adversarial Training and Correcting Stereotype of Noisy Labelscited by 8
2025AdPO: Enhancing the Adversarial Robustness of Large Vision-Language Models with Preference Optimizationcited by 8
2021Self-Supervised Adversarial Example Detection by Disentangled Representationcited by 7
2021The art of defense: letting networks fool the attackercited by 6
2021Selection of Source Images Heavily Influences the Effectiveness of Adversarial Attackscited by 6
2021Adaptive Perturbation for Adversarial Attackcited by 6
2022Special Session: Towards an Agile Design Methodology for Efficient, Reliable, and Secure ML Systemscited by 6
2023Detecting Adversarial Directions in Deep Reinforcement Learning to Make Robust Decisionscited by 6
2023Certifiable Black-Box Attacks with Randomized Adversarial Examples: Breaking Defenses with Provable Confidencecited by 5
2021Simple Post-Training Robustness Using Test Time Augmentations and Random Forestcited by 4
2022Exploring Adversarial Attacks and Defenses in Vision Transformers trained with DINOcited by 4
2022Boosting Adversarial Robustness From The Perspective of Effective Margin Regularizationcited by 4
2024Towards Robust Vision Transformer via Masked Adaptive Ensemblecited by 4
2021Anomaly Detection of Adversarial Examples using Class-conditional Generative Adversarial Networkscited by 3
2021In-distribution adversarial attacks on object recognition models using gradient-free searchcited by 3
2021Unity is strength: Improving the Detection of Adversarial Examples with Ensemble Approachescited by 3
2023Robust Proxy: Improving Adversarial Robustness by Robust Proxy Learningcited by 3
2024Persistent Classification: A New Approach to Stability of Data and Adversarial Examplescited by 3
2024Trustworthy Actionable Perturbationscited by 3
2021BAARD: Blocking Adversarial Examples by Testing for Applicability, Reliability and Decidabilitycited by 2
2021Adversarial Examples Detection with Bayesian Neural Networkcited by 2
2021A Primer on Multi-Neuron Relaxation-based Adversarial Robustness Certificationcited by 2
2021Two Souls in an Adversarial Image: Towards Universal Adversarial Example Detection using Multi-view Inconsistencycited by 2
2022On Trace of PGD-Like Adversarial Attackscited by 2
2022p-DkNN: Out-of-Distribution Detection Through Statistical Testing of Deep Representationscited by 2
2022ADDMU: Detection of Far-Boundary Adversarial Examples with Data and Model Uncertainty Estimationcited by 2
2023Computational Asymmetries in Robust Classificationcited by 2

50 most cited of 75 All 75 (JSON)