mlsanityData to 7 Oct 2026

A Unified Framework for Adversarial Attack and Defense in Constrained Feature Space

arXiv 2112.01156 ↗First posted 2 Dec 2021cs.AI cs.LGcited by 22

Not on the map. Why?

Abstract (arXiv)

The generation of feasible adversarial examples is necessary for properly assessing models that work in constrained feature space. However, it remains a challenging task to enforce constraints into attacks that were designed for computer vision. We propose a unified framework to generate feasible adversarial examples that satisfy given domain constraints. Our framework can handle both linear and non-linear constraints. We instantiate our framework into two algorithms: a gradient-based attack that introduces constraints in the loss function to maximize, and a multi-objective search algorithm that aims for misclassification, perturbation minimization, and constraint satisfaction. We show that our approach is effective in four different domains, with a success rate of up to 100%, where state-of-the-art attacks fail to generate a single feasible example. In addition to adversarial retraining, we propose to introduce engineered non-convex constraints to improve model adversarial robustness. We demonstrate that this new defense is as effective as adversarial retraining. Our framework forms the starting point for research on constrained adversarial attacks and provides relevant baselines and datasets that future research can exploit.

Key facts. Cited by 22 papers in the map (data through 7 Oct 2026). Builds on 21 of its 48 references that are in the map. First posted on arXiv 2 Dec 2021.

Builds on

21 of 48 references.

References · 21

  1. 2013Intriguing properties of neural networkscited by 5,114
  2. 2017Learning Important Features Through Propagating Activation Differencescited by 1,454
  3. 2016Adversarial Machine Learning at Scalecited by 934
  4. 2017Evasion Attacks against Machine Learning at Test Timecited by 686
  5. 2019On Evaluating Adversarial Robustnesscited by 442
  6. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  7. 2018Generating Natural Language Adversarial Examplescited by 422
  8. 2018Adversarial Robustness Toolbox v1.0.0cited by 214
  9. 2019Adversarial Examples for Models of Codecited by 61
  10. 2018Did you hear that? Adversarial Examples Against Automatic Speech Recognitioncited by 48
  11. 2019SirenAttack: Generating Adversarial Audio for End-to-End Acoustic Systemscited by 33
  12. 2019FENCE: Feasible Evasion Attacks on Neural Networks in Constrained Environmentscited by 23
9 more
  1. 2019Physical Adversarial Attacks Against End-to-End Autoencoder Communication Systemscited by 19
  2. 2019Multi-Label Adversarial Perturbationscited by 18
  3. 2021Adversarial Robustness with Non-uniform Perturbationscited by 17
  4. 2021Adversarial Robustness in Multi-Task Learning: Promises and Illusionscited by 17
  5. 2020Towards Benchmark Datasets for Machine Learning Based Website Phishing Detection: An experimental studycited by 16
  6. 2020ConAML: Constrained Adversarial Machine Learning for Cyber-Physical Systemscited by 13
  7. 2020Adversarial Examples in Constrained Domainscited by 11
  8. 2020Crafting Adversarial Examples for Deep Learning Based Prognostics (Extended Version)cited by 9
  9. 2018Evading classifiers in discrete domains with provable optimality guaranteescited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 2; 2024: 7; 2025: 9; 2026: 1.

22 papers.

Sort:

By area:

  1. 2024How Realistic Is Your Synthetic Data? Constraining Deep Generative Models for Tabular Datacited by 29
  2. 2023How do humans perceive adversarial text? A reality check on the validity and naturalness of word-based adversarial attackscited by 18
  3. 2025Beyond the convexity assumption: Realistic tabular data generation under quantifier-free real linear constraintscited by 18
  4. 2022On The Empirical Effectiveness of Unrealistic Adversarial Hardening Against Realistic Adversarial Attackscited by 15
  5. 2024Constrained Adaptive Attack: Effective Adversarial Attack Against Deep Neural Networks for Tabular Datacited by 11
  6. 2024TabularBench: Benchmarking Adversarial Robustness for Tabular Deep Learning in Real-world Use-casescited by 10
  7. 2024Large Language Model Supply Chain: Open Problems From the Security Perspectivecited by 9
  8. 2025Hidden in the Noise: Unveiling Backdoors in Audio LLMs Alignment through Latent Acoustic Pattern Triggerscited by 7
  9. 2024SafePowerGraph: Safety-aware Evaluation of Graph Neural Networks for Transmission Power Gridscited by 5
  10. 2025TabAttackBench: A Benchmark for Adversarial Attacks on Tabular Datacited by 4
  11. 2022On Evaluating Adversarial Robustness of Chest X-ray Classification: Pitfalls and Best Practicescited by 3
  12. 2025On the Robustness of Tabular Foundation Models: Test-Time Attacks and In-Context Defensescited by 3
10 more
  1. 2022Level Up with ML Vulnerability Identification: Leveraging Domain Constraints in Feature Space for Robust Android Malware Detectioncited by 2
  2. 2025Neuro-Symbolic AI for Cybersecurity: State of the Art, Challenges, and Opportunitiescited by 2
  3. 2023Constrained Adaptive Attacks: Realistic Evaluation of Adversarial Examples and Robust Training of Deep Neural Networks for Tabular Datacited by 1
  4. 2024Deep generative models as an adversarial attack strategy for tabular machine learningcited by 1
  5. 2025Crafting Imperceptible On-Manifold Adversarial Attacks for Tabular Datacited by 1
  6. 2024Improving Adversarial Robustness in Android Malware Detection by Reducing the Impact of Spurious Correlationscited by 0
  7. 2025Constraint-Guided Prediction Refinement via Deterministic Diffusion Trajectoriescited by 0
  8. 2025Insights on Adversarial Attacks for Tabular Machine Learning via a Systematic Literature Reviewcited by 0
  9. 2025Constrained Adversarial Perturbationcited by 0
  10. 2026FraudBench: Protocol-Sensitive Benchmarking of Adversarial Robustness for Financial Risk Assessmentcited by 0