mlsanityData to 7 Oct 2026

Robustness of 3D Deep Learning in an Adversarial Setting

arXiv 1904.00923 ↗First posted 1 Apr 2019cs.CV cs.CR cs.LGcited by 55

Not on the map. Why?

Abstract (arXiv)

Understanding the spatial arrangement and nature of real-world objects is of paramount importance to many complex engineering tasks, including autonomous navigation. Deep learning has revolutionized state-of-the-art performance for tasks in 3D environments; however, relatively little is known about the robustness of these approaches in an adversarial setting. The lack of comprehensive analysis makes it difficult to justify deployment of 3D deep learning models in real-world, safety-critical applications. In this work, we develop an algorithm for analysis of pointwise robustness of neural networks that operate on 3D data. We show that current approaches presented for understanding the resilience of state-of-the-art models vastly overestimate their robustness. We then use our algorithm to evaluate an array of state-of-the-art models in order to demonstrate their vulnerability to occlusion attacks. We show that, in the worst case, these networks can be reduced to 0% classification accuracy after the occlusion of at most 6.5% of the occupied input space.

Key facts. Cited by 55 papers in the map (data through 7 Oct 2026). First posted on arXiv 1 Apr 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 13; 2022: 14; 2023: 9; 2024: 7; 2025: 9; 2026: 3.

55 papers.

Sort:

By area:

  1. 20223D Object Detection for Autonomous Driving: A Comprehensive Surveycited by 132
  2. 2021Advances in adversarial attacks and defenses in computer vision: A surveycited by 127
  3. 2021PointGuard: Provably Robust 3D Point Cloud Classificationcited by 49
  4. 2021Imperceptible Transfer Attack and Defense on 3D Point Cloud Classificationcited by 36
  5. 2021PointBA: Towards Backdoor Attacks in 3D Point Cloudcited by 34
  6. 2021Deep Learning-Based Autonomous Driving Systems: A Survey of Attacks and Defensescited by 29
  7. 2021Fooling LiDAR Perception via Adversarial Trajectory Perturbationcited by 25
  8. 2022Isometric 3D Adversarial Examples in the Physical Worldcited by 25
  9. 2022Physical-World Optical Adversarial Attacks on 3D Face Recognitioncited by 23
  10. 2022A Comprehensive Study of the Robustness for LiDAR-based 3D Object Detectors against Adversarial Attackscited by 22
  11. 2022Exploring the Devil in Graph Spectral Domain for 3D Point Cloud Attackscited by 19
  12. 2023A Survey of Robustness and Safety of 2D and 3D Deep Learning Models Against Adversarial Attackscited by 19
38 more
  1. 2022How Do We Fail? Stress Testing Perception in Autonomous Vehiclescited by 17
  2. 2022Robust Structured Declarative Classifiers for 3D Point Clouds: Defending Adversarial Attacks with Implicit Gradientscited by 17
  3. 2021Robustness Certification for Point Cloud Modelscited by 16
  4. 2022Point Cloud Attacks in Graph Spectral Domain: When 3D Geometry Meets Graph Signal Processingcited by 16
  5. 2023From Robustness to Explainability and Back Againcited by 14
  6. 2022OccAM's Laser: Occlusion-based Attribution Maps for 3D Object Detectors on LiDAR Datacited by 13
  7. 2022LPF-Defense: 3D Adversarial Defense based on Frequency Analysiscited by 10
  8. 20233DHacker: Spectrum-based Decision Boundary Generation for Hard-label 3D Point Cloud Attackcited by 9
  9. 2021Local Aggressive Adversarial Attacks on 3D Point Cloudcited by 7
  10. 2021Explainability-Aware One Point Attack for Point Cloud Neural Networkscited by 7
  11. 2022Invariance-Aware Randomized Smoothing Certificatescited by 7
  12. 2021Adversarial Attack by Limited Point Cloud Surface Modificationscited by 6
  13. 2023PointCert: Point Cloud Classification with Deterministic Certified Robustness Guaranteescited by 6
  14. 2023Risk-optimized Outlier Removal for Robust 3D Point Cloud Classificationcited by 6
  15. 2021Understanding and Measuring Robustness of Multimodal Learningcited by 5
  16. 20223DVerifier: Efficient Robustness Verification for 3D Point Cloud Modelscited by 5
  17. 2023Adversarial Attacks and Defenses on 3D Point Cloud Classification: A Surveycited by 5
  18. 2025INTACT: Inducing Noise Tolerance through Adversarial Curriculum Training for LiDAR-based Safety-Critical Perception and Autonomycited by 5
  19. 2021On Adversarial Robustness of Point Cloud Semantic Segmentationcited by 3
  20. 2022Random Walks for Adversarial Meshescited by 3
  21. 2021Generating Unrestricted 3D Adversarial Point Cloudscited by 2
  22. 2023Robustifying Point Cloud Networks by Refocusingcited by 2
  23. 2024PCLD: Point Cloud Layerwise Diffusion for Adversarial Purificationcited by 2
  24. 2024Robust 3D Point Clouds Classification based on Declarative Defenderscited by 2
  25. 2023SHIFT3D: Synthesizing Hard Inputs For Tricking 3D Detectorscited by 1
  26. 2024Flow AM: Generating Point Cloud Global Explanations by Latent Alignmentcited by 1
  27. 2024Eidos: Efficient, Imperceptible Adversarial 3D Point Cloudscited by 1
  28. 2024LiDAttack: Robust Black-box Attack on LiDAR-based Object Detectioncited by 1
  29. 2025Transferable and Undefendable Point Cloud Attacks via Medial Axis Transformcited by 1
  30. 2025Generating Adversarial Point Clouds Using Diffusion Modelcited by 1
  31. 20223D-VFD: A Victim-free Detector against 3D Adversarial Point Cloudscited by 0
  32. 2023Adaptive Local Adversarial Attacks on 3D Point Clouds for Augmented Realitycited by 0
  33. 2024Hard-Label Black-Box Attacks on 3D Point Cloudscited by 0
  34. 2024Imperceptible Adversarial Attacks on Point Clouds Guided by Point-to-Surface Fieldcited by 0
  35. 2025Hi-ALPS -- An Experimental Robustness Quantification of Six LiDAR-based Object Detection Systems for Autonomous Drivingcited by 0
  36. 2025Rethinking Gradient-based Adversarial Attacks on Point Cloud Classificationcited by 0
  37. 2025Cage-Based Deformation for Transferable and Undefendable Point Cloud Attackcited by 0
  38. 2025Robustifying 3D Perception via Least-Squares Graphs for Multi-Agent Object Trackingcited by 0

50 most cited of 55 All 55 (JSON)