mlsanityData to 7 Oct 2026

MACER: Attack-free and Scalable Robust Training via Maximizing Certified Radius

arXiv 2001.02378 ↗First posted 8 Jan 2020cs.LG cs.CR stat.MLcited by 114

Not on the map. Why?

Abstract (arXiv)

Adversarial training is one of the most popular ways to learn robust models but is usually attack-dependent and time costly. In this paper, we propose the MACER algorithm, which learns robust models without using adversarial training but performs better than all existing provable l2-defenses. Recent work shows that randomized smoothing can be used to provide a certified l2 radius to smoothed classifiers, and our algorithm trains provably robust smoothed classifiers via MAximizing the CErtified Radius (MACER). The attack-free characteristic makes MACER faster to train and easier to optimize. In our experiments, we show that our method can be applied to modern deep neural networks on a wide range of datasets, including Cifar-10, ImageNet, MNIST, and SVHN. For all tasks, MACER spends less training time than state-of-the-art adversarial training algorithms, and the learned models achieve larger average certified radius.

Key facts. Cited by 114 papers in the map (data through 7 Oct 2026). First posted on arXiv 8 Jan 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 22; 2022: 28; 2023: 23; 2024: 24; 2025: 8; 2026: 9.

114 papers.

Sort:

By area:

  1. 2022(Certified!!) Adversarial Robustness for Free!cited by 130
  2. 2021Advances in adversarial attacks and defenses in computer vision: A surveycited by 127
  3. 2021Globally-Robust Neural Networkscited by 106
  4. 2021SmoothMix: Training Confidence-calibrated Smoothed Classifiers for Certified Robustnesscited by 49
  5. 2022Rethinking Lipschitz Neural Networks and Certified Robustness: A Boolean Function Perspectivecited by 48
  6. 2021Boosting Randomized Smoothing with Variance Reduced Classifierscited by 47
  7. 2021On the Certified Robustness for Ensemble Models and Beyondcited by 40
  8. 2021Improved, Deterministic Smoothing for L_1 Certified Robustnesscited by 39
  9. 2021Scalable Certified Segmentation via Randomized Smoothingcited by 35
  10. 2022DensePure: Understanding Diffusion Models towards Adversarial Robustnesscited by 33
  11. 2021Enhancing Adversarial Robustness via Test-time Transformation Ensemblingcited by 27
  12. 2024Provably Robust Conformal Prediction with Improved Efficiencycited by 27
38 more
  1. 2024Large Language Models and Causal Inference in Collaboration: A Surveycited by 26
  2. 2021Knowledge Enhanced Machine Learning Pipeline against Diverse Adversarial Attackscited by 24
  3. 2021Boosting the Certified Robustness of L-infinity Distance Netscited by 24
  4. 2024Your Diffusion Model is Secretly a Certifiably Robust Classifiercited by 24
  5. 2021ANCER: Anisotropic Certification via Sample-wise Volume Maximizationcited by 23
  6. 2022Learning Physics-Informed Neural Networks without Stacked Back-propagationcited by 23
  7. 2022Certified Robustness Against Natural Language Attacks by Causal Interventioncited by 23
  8. 2022A Survey of Adversarial Defences and Robustness in NLPcited by 22
  9. 2022Double Sampling Randomized Smoothingcited by 22
  10. 2022Denoising Masked AutoEncoders Help Robust Classificationcited by 21
  11. 2023DiffSmooth: Certifiably Robust Learning via Diffusion Models and Local Smoothingcited by 20
  12. 2021CC-Cert: A Probabilistic Approach to Certify General Robustness of Neural Networkscited by 18
  13. 2024Provably Unlearnable Data Examplescited by 18
  14. 2021DeformRS: Certifying Input Deformations with Randomized Smoothingcited by 17
  15. 2021Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselinescited by 17
  16. 2022Double Bubble, Toil and Trouble: Enhancing Certified Robustness through Transitivitycited by 15
  17. 2024Mitigating the Curse of Dimensionality for Certified Robustness via Dual Randomized Smoothingcited by 15
  18. 2022Robust and Accurate -- Compositional Architectures for Randomized Smoothingcited by 12
  19. 2021Input-Specific Robustness Certification for Randomized Smoothingcited by 11
  20. 2022Input-agnostic Certified Group Fairness via Gaussian Parameter Smoothingcited by 11
  21. 2023Projected Randomized Smoothing for Certified Adversarial Robustnesscited by 11
  22. 20223DeformRS: Certifying Spatial Deformations on Point Cloudscited by 10
  23. 2022CARE: Certifiably Robust Learning with Reasoning via Variational Inferencecited by 10
  24. 2022Towards Robust Dataset Learningcited by 9
  25. 2022Confidence-aware Training of Smoothed Classifiers for Certified Robustnesscited by 9
  26. 2023Multi-scale Diffusion Denoised Smoothingcited by 9
  27. 2023Asymmetric Certified Robustness via Feature-Convex Neural Networkscited by 8
  28. 2023Node-aware Bi-smoothing: Certified Robustness against Graph Injection Attackscited by 8
  29. 2022Accelerating Certified Robustness Training via Knowledge Transfercited by 7
  30. 2023Understanding Noise-Augmented Training for Randomized Smoothingcited by 7
  31. 2023Improving Viewpoint Robustness for Visual Recognition via Adversarial Trainingcited by 7
  32. 2023Improve Robustness of Reinforcement Learning against Observation Perturbations via l_∞ Lipschitz Policy Networkscited by 7
  33. 2022Generalizability of Adversarial Robustness Under Distribution Shiftscited by 6
  34. 2023Augment then Smooth: Reconciling Differential Privacy with Certified Robustnesscited by 6
  35. 2021Insta-RS: Instance-wise Randomized Smoothing for Improved Robustness and Accuracycited by 5
  36. 2022(De-)Randomized Smoothing for Decision Stump Ensemblescited by 5
  37. 2022Data Models for Dataset Drift Controls in Machine Learning With Optical Imagescited by 5
  38. 2023A Certified Radius-Guided Attack Framework to Image Segmentation Modelscited by 5

50 most cited of 114 All 114 (JSON)