mlsanityData to 7 Oct 2026

CausalDiff: Causality-Inspired Disentanglement via Diffusion Model for Adversarial Defense

arXiv 2410.23091 ↗First posted 30 Oct 2024cs.CVcited by 4

On the map: Diffusion models → Robustness, privacy and adversarial aspects

Abstract (arXiv)

Despite ongoing efforts to defend neural classifiers from adversarial attacks, they remain vulnerable, especially to unseen attacks. In contrast, humans are difficult to be cheated by subtle manipulations, since we make judgments only based on essential factors. Inspired by this observation, we attempt to model label generation with essential label-causative factors and incorporate label-non-causative factors to assist data generation. For an adversarial example, we aim to discriminate the perturbations as non-causative factors and make predictions only based on the label-causative factors. Concretely, we propose a casual diffusion model (CausalDiff) that adapts diffusion models for conditional data generation and disentangles the two types of casual factors by learning towards a novel casual information bottleneck objective. Empirically, CausalDiff has significantly outperformed state-of-the-art defense methods on various unseen attacks, achieving an average robustness of 86.39% (+4.01%) on CIFAR-10, 56.25% (+3.13%) on CIFAR-100, and 82.62% (+4.93%) on GTSRB (German Traffic Sign Recognition Benchmark). The code is available at https://github.com/CAS-AISafetyBasicResearchGroup/CausalDiff.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in Robustness, privacy and adversarial aspects, one of 14 directions in Diffusion models. Builds on 50 of its 66 references that are in the map. First posted on arXiv 30 Oct 2024.

Builds on

50 of 66 references.

Closest ancestors · 7

Same direction, newest first

  1. 2023Robust Classification via a Single Diffusion Model24 May
  2. 2023Better Diffusion Models Further Improve Adversarial Training9 Feb
  3. 2022DensePure: Understanding Diffusion Models towards Adversarial Robustness1 Nov
  4. 2022Guided Diffusion Model for Adversarial Purification from Random Noise22 Jun
  5. 2022Guided Diffusion Model for Adversarial Purification30 May
  6. 2022Diffusion Models for Adversarial Purification16 May
  7. 2021Adversarial purification with Score-based generative models11 Jun

Other references · 43

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  5. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  6. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  7. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,539
  8. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  9. 2022Classifier-Free Diffusion Guidancecited by 5,717
  10. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  11. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,595
  12. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 2,994
31 more
  1. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  2. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  3. 2019Adversarial Training for Free!cited by 638
  4. 2021Diffusion Autoencoders: Toward a Meaningful and Decodable Representationcited by 447
  5. 2018Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Modelscited by 362
  6. 2019SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimizationcited by 253
  7. 2019Causality for Machine Learningcited by 242
  8. 2020CLUB: A Contrastive Log-ratio Upper Bound of Mutual Informationcited by 233
  9. 2021Improving Robustness using Generated Datacited by 233
  10. 2021Fixing Data Augmentation to Improve Adversarial Robustnesscited by 201
  11. 2020Representation Learning via Invariant Causal Mechanismscited by 197
  12. 2018Spatially Transformed Adversarial Examplescited by 195
  13. 2017Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulationcited by 182
  14. 2018MMA Training: Direct Input Space Margin Maximization through Adversarial Trainingcited by 171
  15. 2019Adversarial Robustness through Local Linearizationcited by 146
  16. 2022Learning Causally Invariant Representations for Out-of-Distribution Generalization on Graphscited by 144
  17. 2020Perceptual Adversarial Robustness: Defense Against Unseen Threat Modelscited by 137
  18. 2020A unifying mutual information view of metric learning: cross-entropy vs. pairwise lossescited by 100
  19. 2020Causal Interpretability for Machine Learning -- Problems, Methods and Evaluationcited by 82
  20. 2019Defending Against Physically Realizable Attacks on Image Classificationcited by 66
  21. 2020Learning Causal Semantic Representation for Out-of-Distribution Predictioncited by 60
  22. 2020Stochastic Security: Adversarial Defense Using Long-Run Dynamics of Energy-Based Modelscited by 58
  23. 2020A Causal View on Robustness of Neural Networkscited by 53
  24. 2021Score-Based Generative Classifierscited by 53
  25. 2022Out-of-distribution Generalization with Causal Invariant Transformationscited by 45
  26. 2021CausalAdv: Adversarial Robustness through the Lens of Causalitycited by 37
  27. 2020Causality Learning: A New Perspective for Interpretable Machine Learningcited by 28
  28. 2023Demystifying Causal Features on Adversarial Examples and Causal Inoculation for Robust Network by Adversarial Instrumental Variable Regressioncited by 18
  29. 2021Adversarial Visual Robustness by Causal Interventioncited by 13
  30. 2023Mitigating Adversarial Vulnerability through Causal Parameter Estimation by Adversarial Double Machine Learningcited by 11
  31. 2022Causal Information Bottleneck Boosts Adversarial Robustness of Deep Neural Networkcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2025CLIPure: Purification in Latent Space via CLIP for Adversarially Robust Zero-Shot Classificationcited by 12
  2. 2025NAPPure: Adversarial Purification for Robust Image Classification under Non-Additive Perturbationscited by 0
  3. 2026MuCALD-SplitFed: Causal-Latent Diffusion for Privacy-Preserving Multi-Task Split-Federated Medical Image Segmentationcited by 0
  4. 2026Clinically-Grounded Counterfactual Reasoning for Medical Video Diagnosiscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Diffusion or Non-Diffusion Adversarial Defenses: Rethinking the Relation between Classifier and Adversarial Purifier
  2. 2023DiffDefense: Defending against Adversarial Attacks via Diffusion Models
  3. 2025DiffSDA: Unsupervised Diffusion Sequential Disentanglement Across Modalities
  4. 2024DiffBreak: Is Diffusion-Based Purification Robust?
  5. 2024CausAdv: A Causal-based Framework for Detecting Adversarial Examples
  6. 2024Watch the Watcher! Backdoor Attacks on Security-Enhancing Diffusion Models
  7. 2024Causal Diffusion Autoencoders: Toward Counterfactual Generation via Diffusion Probabilistic Models
  8. 2023DIFFender: Diffusion-Based Adversarial Defense against Patch Attacks