Research map / Point clouds / Point cloud classification
Shape Prior Guided Attack: Sparser Perturbations on 3D Point Clouds
Zhenbo Shi, Zhi Chen, Zhenbo Xu, Wei Yang, Zhidong Yu, Liusheng Huang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use SPGA instead of AdvPC | Point cloud classifier attacks · ModelNet40, defended classifiers | attack success rate | Improved |
| Use SPGA rather than the strongest listed baseline | Attacks on PointNet with RO defense · ModelNet40 | attack success rate: 86.8 -> 87.5 | Improved |
| Use SPGA rather than the strongest listed baseline | Attacks on PointNet with RS defense · ModelNet40 | attack success rate: 89.5 -> 88.6 | Hurt |
| Add FOFA optimization (ablation) | SPGA · Adversarial sample generation | generation time: 24.47 -> 18.09 | Improved |
| Add FOFA optimization (ablation) | SPGA · ModelNet40, untargeted attack | perturbation budget (%): 4.82 -> 4.36 | Improved |
| Allow attaching perturbation points only (SPGA-AT) instead of full SPGA (ablation) | SPGA · ModelNet40, defended classifiers | attack success rate | Mixed |
| Allow dropping original points only (SPGA-DR) instead of full SPGA (ablation) | SPGA · ModelNet40, defended classifiers | attack success rate | Mixed |
| Add FOFA optimization (ablation) | SPGA-AT · Adversarial sample generation | generation time: 24.47 -> 18.63 | Improved |
| Add FOFA optimization (ablation) | SPGA-DR · Adversarial sample generation | generation time: 24.03 -> 18.15 | Improved |
| Use shape-prior-guided sparse perturbations | SPGA · ModelNet40, untargeted attack | perturbation budget (%): 7.64 -> 4.36 | Improved |
| Use shape-prior-guided sparse perturbations | SPGA · ModelNet40, untargeted attack | perturbation budget (%): 5.29 -> 4.36 | Improved |
| Increase the number of SPGA adversarial points | SPGA · 3D MNIST | classification accuracy | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AdvPC — AdvPC: Transferable Adversarial Perturbations on 3D Point Clouds (ECCV 2020) · better on 8 of 8 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Hide in Thicket: Generating Imperceptible and Rational Adversarial Perturbations on 3D Point Clouds (CVPR 2024 · similarity 0.91)
- Generating 3D Adversarial Point Clouds (CVPR 2019 · similarity 0.90)
- Transferable 3D Adversarial Shape Completion using Diffusion Models (ECCV 2024 · similarity 0.90)
- Explicitly Perceiving and Preserving the Local Geometric Structures for 3D Point Cloud Attack (AAAI 2024 · similarity 0.90)
- Curvature-Invariant Adversarial Attacks for 3D Point Clouds (AAAI 2024 · similarity 0.88)
- Manifold Constraints for Imperceptible Adversarial Attacks on Point Clouds (AAAI 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
