mlsanityData to 7 Oct 2026

Universal Spectral Adversarial Attacks for Deformable Shapes

arXiv 2104.03356 ↗First posted 7 Apr 2021cs.LGcited by 14

Not on the map. Why?

Abstract (arXiv)

Machine learning models are known to be vulnerable to adversarial attacks, namely perturbations of the data that lead to wrong predictions despite being imperceptible. However, the existence of "universal" attacks (i.e., unique perturbations that transfer across different data points) has only been demonstrated for images to date. Part of the reason lies in the lack of a common domain, for geometric data such as graphs, meshes, and point clouds, where a universal perturbation can be defined. In this paper, we offer a change in perspective and demonstrate the existence of universal attacks for geometric data (shapes). We introduce a computational procedure that operates entirely in the spectral domain, where the attacks take the form of small perturbations to short eigenvalue sequences; the resulting geometry is then synthesized via shape-from-spectrum recovery. Our attacks are universal, in that they transfer across different shapes, different representations (meshes and point clouds), and generalize to previously unseen data.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). Builds on 49 of its 66 references that are in the map. First posted on arXiv 7 Apr 2021.

Builds on

49 of 66 references.

References · 49

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  4. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  5. 2013Intriguing properties of neural networkscited by 5,114
  6. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  7. 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,085
  8. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  9. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,259
  10. 2016Practical Black-Box Attacks against Machine Learningcited by 893
  11. 2016Universal adversarial perturbationscited by 774
  12. 2017ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Modelscited by 656
37 more
  1. 2016Delving into Transferable Adversarial Examples and Black-box Attackscited by 530
  2. 2017Adversarial Patchcited by 514
  3. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  4. 2018Adversarial Attacks on Neural Networks for Graph Datacited by 395
  5. 2018Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifierscited by 341
  6. 2018Feature Denoising for Improving Adversarial Robustnesscited by 312
  7. 20163D Menagerie: Modeling the 3D shape and pose of animalscited by 310
  8. 2018Adversarial Attack on Graph Structured Datacited by 300
  9. 2018Generating 3D faces using Convolutional Mesh Autoencoderscited by 267
  10. 2019Adversarial Attacks and Defenses in Images, Graphs and Text: A Reviewcited by 234
  11. 2019Adversarial Policies: Attacking Deep Reinforcement Learningcited by 209
  12. 2017Generative Adversarial Perturbationscited by 153
  13. 2018Generating 3D Adversarial Point Cloudscited by 144
  14. 2018Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness against Adversarial Attackcited by 123
  15. 2018Decoupling Direction and Norm for Efficient Gradient-Based L2 Adversarial Attacks and Defensescited by 109
  16. 2019NATTACK: Learning the Distributions of Adversarial Examples for an Improved Black-Box Attack on Deep Neural Networkscited by 94
  17. 2017Universal Adversarial Perturbations Against Semantic Image Segmentationcited by 92
  18. 2019Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud Classifierscited by 92
  19. 2018Generalizable Data-free Objective for Crafting Universal Adversarial Perturbationscited by 88
  20. 2018Universal Adversarial Trainingcited by 85
  21. 2019AdvPC: Transferable Adversarial Perturbations on 3D Point Cloudscited by 84
  22. 2017NAG: Network for Adversary Generationcited by 51
  23. 2020On Isometry Robustness of Deep 3D Point Cloud Models under Adversarial Attackscited by 46
  24. 2019Data Poisoning Attack against Knowledge Graph Embeddingcited by 34
  25. 2019Defending against Whitebox Adversarial Attacks via Randomized Discretizationcited by 29
  26. 2020Universal Adversarial Perturbations: A Surveycited by 28
  27. 2018Data Poisoning Attack against Unsupervised Node Embedding Methodscited by 19
  28. 2018Isospectralization, or how to hear shape, style, and correspondencecited by 18
  29. 2020Graph Universal Adversarial Attacks: A Few Bad Actors Ruin Graph Learning Modelscited by 18
  30. 2014On the optimality of shape and data representation in the spectral domaincited by 14
  31. 2020Defense-PointNet: Protecting PointNet Against Adversarial Attackscited by 13
  32. 2020Instant recovery of shape from spectrum via latent space connectionscited by 11
  33. 2017Practical Attacks Against Graph-based Clusteringcited by 9
  34. 2019Geometry-Aware Generation of Adversarial Point Cloudscited by 6
  35. 2019Exploring the Robustness of NMT Systems to Nonsensical Inputscited by 5
  36. 2019Adversarial Training with Voronoi Constraintscited by 4
  37. 2019Enforcing Linearity in DNN succours Robustness and Adversarial Image Generationcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 4; 2022: 3; 2023: 2; 2024: 2; 2025: 2; 2026: 1.

14 papers.

Sort:

By area:

  1. 2021Advances in adversarial attacks and defenses in computer vision: A surveycited by 127
  2. 2023Enhancing Generalization of Universal Adversarial Perturbation through Gradient Aggregationcited by 22
  3. 2021Stereoscopic Universal Perturbations across Different Architectures and Datasetscited by 17
  4. 2021Graph Structural Attack by Perturbing Spectral Distancecited by 16
  5. 2022Scattering Model Guided Adversarial Examples for SAR Target Recognition: Attack and Defensecited by 14
  6. 2024Generating Graphs via Spectral Diffusioncited by 8
  7. 2023Latent Spectral Regularization for Continual Learningcited by 6
  8. 2025ETA: Energy-based Test-time Adaptation for Depth Completioncited by 4
  9. 2022Random Walks for Adversarial Meshescited by 3
  10. 2022SAGA: Spectral Adversarial Geometric Attack on 3D Meshescited by 2
  11. 2024DifAttack++: Query-Efficient Black-Box Adversarial Attack via Hierarchical Disentangled Feature Space in Cross-Domaincited by 2
  12. 2025NoPain: No-box Point Cloud Attack via Optimal Transport Singular Boundarycited by 2
2 more
  1. 2021Localized Shape Modelling with Global Coherence: An Inverse Spectral Approachcited by 0
  2. 2026MME: Mixture of Mesh Experts with Random Walk Transformer Gatingcited by 0