Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
126 tokens/sec
GPT-4o
47 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

PCLD: Point Cloud Layerwise Diffusion for Adversarial Purification (2403.06698v1)

Published 11 Mar 2024 in cs.CV

Abstract: Point clouds are extensively employed in a variety of real-world applications such as robotics, autonomous driving and augmented reality. Despite the recent success of point cloud neural networks, especially for safety-critical tasks, it is essential to also ensure the robustness of the model. A typical way to assess a model's robustness is through adversarial attacks, where test-time examples are generated based on gradients to deceive the model. While many different defense mechanisms are studied in 2D, studies on 3D point clouds have been relatively limited in the academic field. Inspired from PointDP, which denoises the network inputs by diffusion, we propose Point Cloud Layerwise Diffusion (PCLD), a layerwise diffusion based 3D point cloud defense strategy. Unlike PointDP, we propagated the diffusion denoising after each layer to incrementally enhance the results. We apply our defense method to different types of commonly used point cloud models and adversarial attacks to evaluate its robustness. Our experiments demonstrate that the proposed defense method achieved results that are comparable to or surpass those of existing methodologies, establishing robustness through a novel technique. Code is available at https://github.com/batuceng/diffusion-layer-robustness-pc.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (43)
  1. “Pointnet: Deep learning on point sets for 3d classification and segmentation,” in IEEE/CVF CVPR, July 2017.
  2. “Escape from cells: Deep kd-networks for the recognition of 3d point cloud models,” in ICCV, 2017, pp. 863–872.
  3. “Fat: Field-aware transformer for 3d point cloud semantic segmentation,” in IEEE ICIP. IEEE, 2023, pp. 660–664.
  4. “Laptran: Transformer embedding graph laplacian for point cloud part segmentation,” in IEEE ICIP. IEEE, 2023, pp. 3070–3074.
  5. “Ppf-foldnet: Unsupervised learning of rotation invariant 3d local descriptors,” in ECCV, 2018, pp. 602–618.
  6. “Partial point cloud registration via soft segmentation,” in IEEE ICIP. IEEE, 2022, pp. 681–685.
  7. “Fully automated scan-to-bim via point cloud instance segmentation,” in IEEE ICIP. IEEE, 2023, pp. 291–295.
  8. “Deep learning approach to point cloud scene understanding for automated scan to 3d reconstruction,” Journal of Computing in Civil Engineering, vol. 33, no. 4, pp. 04019027, 2019.
  9. “Pct: Large-scale 3d point cloud representations via graph inception networks with applications to autonomous driving,” in IEEE ICIP. IEEE, 2019, pp. 4395–4399.
  10. “4d panoptic lidar segmentation,” in IEEE/CVF CVPR, 2021, pp. 5527–5537.
  11. “Adversarial sensor attack on lidar-based perception in autonomous driving,” in ACM SIGSAC conference on computer and communications security, 2019, pp. 2267–2281.
  12. “High-precision point cloud registration system of multi-view industrial self-similar workpiece based on super-point space guidance,” Journal of Intelligent Manufacturing, pp. 1–15, 2023.
  13. “Review of artificial intelligence adversarial attack and defense technologies,” Applied Sciences, vol. 9, no. 5, pp. 909, 2019.
  14. “Towards deep learning models resistant to adversarial attacks,” arXiv:1706.06083, 2017.
  15. “Diffusion models for adversarial purification,” arXiv:2205.07460, 2022.
  16. “Robustness of 3d deep learning in an adversarial setting,” in IEEE/CVF CVPR, 2019, pp. 11767–11775.
  17. “Generating 3d adversarial point clouds,” in IEEE/CVF CVPR, 2019, pp. 9136–9144.
  18. “Adversarially robust 3d point cloud recognition using self-supervisions,” NeurIPS, vol. 34, pp. 15498–15512, 2021.
  19. “Dup-net: Denoiser and upsampler network for 3d adversarial point clouds defense,” in IEEE/CVF CVPR, 2019, pp. 1961–1970.
  20. “If-defense: 3d adversarial point cloud defense via implicit function based restoration,” arXiv preprint arXiv:2010.05272, 2020.
  21. “Improving adversarial robustness of 3d point cloud classification models,” in ECCV. Springer, 2022, pp. 672–689.
  22. “Dynamic graph cnn for learning on point clouds,” ACM TOG, vol. 38, no. 5, pp. 1–12, 2019.
  23. “Pointnet++: Deep hierarchical feature learning on point sets in a metric space,” NeurIPS, vol. 30, 2017.
  24. “Pct: Point cloud transformer,” Computational Visual Media, vol. 7, pp. 187–199, 2021.
  25. “Rethinking network design and local geometry in point cloud: A simple residual mlp framework,” arXiv:2202.07123, 2022.
  26. “Curvenet: Curvature-based multitask learning deep networks for 3d object recognition,” IEEE/CAA Journal of Automatica Sinica, vol. 8, no. 6, pp. 1177–1187, 2021.
  27. “Intriguing properties of neural networks,” arXiv:1312.6199, 2013.
  28. “Explaining and harnessing adversarial examples,” arXiv:1412.6572, 2014.
  29. “Adversarial examples in the physical world,” in Artificial intelligence safety and security, pp. 99–112. Chapman and Hall/CRC, 2018.
  30. “Towards evaluating the robustness of neural networks,” in 2017 ieee symposium on security and privacy (sp). Ieee, 2017, pp. 39–57.
  31. “Extending adversarial attacks and defenses to deep 3d point cloud classifiers,” in IEEE ICIP. IEEE, 2019, pp. 2279–2283.
  32. “Pointcloud saliency maps,” in IEEE/CVF CVPR, 2019, pp. 1598–1606.
  33. “Robust adversarial objects against deep learning models,” in AAAI, 2020, vol. 34, pp. 954–962.
  34. “Physically realizable adversarial examples for lidar object detection,” in IEEE/CVF CVPR, June 2020.
  35. “Pointcutmix: Regularization strategy for point cloud classification,” Neurocomputing, vol. 505, pp. 58–67, 2022.
  36. “Online adversarial purification based on self-supervision,” arXiv:2101.09387, 2021.
  37. “Denoising diffusion probabilistic models,” NeurIPS, vol. 33, pp. 6840–6851, 2020.
  38. “Towards 3d point cloud based object maps for household environments,” Robotics and Autonomous Systems, vol. 56, no. 11, pp. 927–941, 2008.
  39. “Pu-net: Point cloud upsampling network,” in IEEE/CVF CVPR, 2018, pp. 2790–2799.
  40. “Occupancy networks: Learning 3d reconstruction in function space,” in IEEE/CVF CVPR, 2019, pp. 4460–4470.
  41. “A critical revisit of adversarial robustness in 3d point cloud recognition with diffusion-driven purification,” in ICML. PMLR, 2023, pp. 33100–33114.
  42. “Diffusion probabilistic models for 3d point cloud generation,” CoRR, vol. abs/2103.01458, 2021.
  43. “3d shapenets: A deep representation for volumetric shapes,” in IEEE/CVF CVPR, 2015, pp. 1912–1920.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now