Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/cuge1995/awesome-3D-point-cloud-attacks

List of state of the art papers, code, and other resources
https://github.com/cuge1995/awesome-3D-point-cloud-attacks

List: awesome-3D-point-cloud-attacks

adversarial-attacks adversarial-example adversarial-point-clouds attacks deep-learning defense point-cloud

Last synced: about 1 month ago
JSON representation

List of state of the art papers, code, and other resources

Awesome Lists containing this project

README 

        
# awesome-3D-point-cloud-attacks

List of state of the art papers, code, and other resources focus on 3D point cloud attacks and defense



## Attacks



- [Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud Classifiers.](https://arxiv.org/pdf/1901.03006.pdf) 

  - [[Code](https://github.com/Daniel-Liu-c0deb0t/3D-Neural-Network-Adversarial-Attacks)]



- [Adversarial point perturbations on 3D objects.](https://arxiv.org/pdf/1908.06062.pdf)   `ECCV 2020`

  - [[Code](https://github.com/Daniel-Liu-c0deb0t/Adversarial-point-perturbations-on-3D-objects)]



- [PointCloud Saliency Maps.](https://openaccess.thecvf.com/content_ICCV_2019/papers/Zheng_PointCloud_Saliency_Maps_ICCV_2019_paper.pdf)  `ICCV 2019` `drop points attack`

  - [[Code](https://github.com/tianzheng4/Learning-PointCloud-Saliency-Maps)]



- [LG-GAN: Label Guided Adversarial Network for Flexible Targeted Attack of Point Cloud Based Deep Networks.](https://openaccess.thecvf.com/content_CVPR_2020/html/Zhou_LG-GAN_Label_Guided_Adversarial_Network_for_Flexible_Targeted_Attack_of_CVPR_2020_paper.html)  `CVPR 2020`

  - [[Code](https://github.com/RyanHangZhou/LG-GAN)]



- [Robust Adversarial Objects against Deep Learning Models.](https://www.aaai.org/ojs/index.php/AAAI/article/view/5443/5299)  `AAAI 2020` `knn attack`

  - [[Code](https://github.com/jinyier/ai_pointnet_attack)]



- [Self-Robust 3D Point Recognition via Gather-Vector Guidance.](https://openaccess.thecvf.com/content_CVPR_2020/papers/Dong_Self-Robust_3D_Point_Recognition_via_Gather-Vector_Guidance_CVPR_2020_paper.pdf)  `CVPR 2020` `FGM IFGM MIFGM PGD`



- [On Isometry Robustness of Deep 3D Point Cloud Models Under Adversarial Attacks.](http://openaccess.thecvf.com/content_CVPR_2020/papers/Zhao_On_Isometry_Robustness_of_Deep_3D_Point_Cloud_Models_Under_CVPR_2020_paper.pdf) `CVPR 2020`

  - [[Code](https://github.com/skywalker6174/3d-isometry-robust)] 



- [Adversarial Autoencoders for Generating 3D Point Clouds.](https://arxiv.org/abs/1811.07605.pdf)  



- [Generating 3D Adversarial Point Clouds.](https://arxiv.org/abs/1809.07016.pdf)  `CVPR 2019` `CW attack`

  - [[Code](https://github.com/xiangchong1/3d-adv-pc)] 



- [PU-GAN: a Point Cloud Upsampling Adversarial Network.](https://arxiv.org/pdf/1907.10844.pdf)  

  - [[Code](https://github.com/liruihui/PU-GAN)]



- [3D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutions.](https://arxiv.org/pdf/1905.06292.pdf)  

  - [[Code](https://github.com/seowok/TreeGAN)]



- [Adversarial Autoencoders for Compact Representations of 3D Point Clouds.](https://arxiv.org/abs/1811.07605v3)  

  - [[Code](https://github.com/MaciejZamorski/3d-AAE)]



- [Physically Realizable Adversarial Examples for LiDAR Object Detection.](https://arxiv.org/pdf/2004.00543.pdf) `CVPR 2020`  



- [AdvPC: Transferable Adversarial Perturbations on 3D Point Clouds.](https://arxiv.org/abs/1912.00461)  `ECCV 2020`

  - [[Code](https://github.com/ajhamdi/AdvPC)]



- [ShapeAdv: Generating Shape-Aware Adversarial 3D Point Clouds.](https://arxiv.org/abs/2005.11626)  

  - [[Code]()]



- [Robustness of 3D Deep Learning in an Adversarial Setting.](https://arxiv.org/abs/1904.00923)  `CVPR 2019`

  - [[Code](https://github.com/matthewwicker/IterativeSalienceOcclusion)]



- [Nudge Attacks on Point-Cloud DNNs.](https://arxiv.org/pdf/2011.11637)



- [Generating Adversarial Surfaces via Band‐Limited Perturbations](https://onlinelibrary.wiley.com/doi/abs/10.1111/cgf.14083)  



- [Geometry-aware generation of adversarial point clouds.](https://arxiv.org/pdf/1912.11171)  `TRAMI`

  - [[Code](https://github.com/Yuxin-Wen/GeoA3)]



- [Efficient Joint Gradient Based Attack Against SOR Defense for 3D Point Cloud Classification.](https://dl.acm.org/doi/abs/10.1145/3394171.3413875)  `ACM MM`



- [Adversarial Objects Against LiDAR Based Autonomous Driving Systems.](https://arxiv.org/pdf/1907.05418.pdf)



- [Minimal Adversarial Examples for Deep Learning on 3D Point Clouds.](https://arxiv.org/pdf/2008.12066.pdf)



- [Towards Robust LiDAR-based Perception in Autonomous Driving: General Black-box Adversarial Sensor Attack and Countermeasures.](https://www.usenix.org/system/files/sec20-sun.pdf)



- [Geometric Adversarial Attacks and Defenses on 3D Point Clouds.](https://arxiv.org/pdf/2012.05657)  `autoencoder attack`

  - [[Code](https://github.com/itailang/geometric_adv)]



- [Towards Universal Physical Attacks On Cascaded Camera-Lidar 3D Object Detection Models.](https://arxiv.org/pdf/2101.10747)



- [Object Removal Attacks on LiDAR-based 3D Object Detectors.](https://arxiv.org/pdf/2102.03722)



- [On the Adversarial Robustness of 3D Point Cloud Classification.](https://arxiv.org/pdf/2011.11922)



- [Exploring Adversarial Robustness of Multi-Sensor Perception Systems in Self Driving.](https://arxiv.org/pdf/2101.06784)



- [Fooling LiDAR Perception via Adversarial Trajectory Perturbation.](https://arxiv.org/pdf/2103.15326.pdf)

  - [[Code](https://ai4ce.github.io/FLAT/)]



- [PointBA: Towards Backdoor Attacks in 3D Point Cloud.](https://arxiv.org/pdf/2103.16074.pdf)



- [3D Adversarial Attacks Beyond Point Cloud.](https://arxiv.org/abs/2104.12146.pdf) `mesh attack`

  - [[Code](https://github.com/cuge1995/Mesh-Attack)]



- [Explainability-Aware One Point Attack for Point Cloud Neural Networks.](https://arxiv.org/abs/2110.04158) `one point attack`

  - [[Code](https://github.com/Explain3D/Exp-One-Point-Atk-PC)]



- [Adversarial Attack by Limited Point Cloud Surface Modifications.](https://arxiv.org/abs/2110.03745)



- [PointBA: Towards Backdoor Attacks in 3D Point Cloud.](https://openaccess.thecvf.com/content/ICCV2021/papers/Li_PointBA_Towards_Backdoor_Attacks_in_3D_Point_Cloud_ICCV_2021_paper.pdf) `ICCV 2021`



- [A Backdoor Attack Against 3D Point Cloud Classifiers.](https://openaccess.thecvf.com/content/ICCV2021/papers/Xiang_A_Backdoor_Attack_Against_3D_Point_Cloud_Classifiers_ICCV_2021_paper.pdf) `ICCV 2021`

 



- [Generating Unrestricted 3D Adversarial Point Clouds.](https://arxiv.org/pdf/2111.08973.pdf)



- [Local Aggressive Adversarial Attack of 3D point Cloud.](https://arxiv.org/pdf/2105.09090.pdf) `ACML oral`

  - [[Code](https://github.com/Chenfeng1271/L3A)]



- [Imperceptible Transfer Attack and Defense on 3D Point Cloud Classification.](https://arxiv.org/pdf/2111.10990.pdf)



- [Attacking Point Cloud Segmentation with Color-only Perturbation](https://arxiv.org/pdf/2112.05871.pdf) `segmentation`



- [Dynamics-aware Adversarial Attack of 3D Sparse Convolution Network](https://arxiv.org/pdf/2112.09428.pdf) `segmentation`



- [Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attacks.](https://arxiv.org/abs/2106.09249) `IEEE S&P`

  - [[Code](https://github.com/ASGuard-UCI/MSF-ADV)]



- [Boosting 3D Adversarial Attacks with Attacking On Frequency](https://arxiv.org/pdf/2201.10937) `AOF attack`

  - [[Code](https://github.com/code-roamer/AOF)]



- [Shape-invariant 3D Adversarial Point Clouds.](https://arxiv.org/pdf/2203.04041.pdf) `CVPR 2022`

  - [[Code](https://github.com/shikiw/SI-Adv)]



- [Shape Prior Guided Attack: Sparser Perturbations on 3D Point Clouds]() `AAAI 2022`



- [Improving transferability of 3D adversarial attacks with scale and shear transformations]() `SS attack`



- [Isometric 3D Adversarial Examples in the Physical World]() `NeurIPS 2022`



- [Rethinking Perturbation Directions for Imperceptible Adversarial Attacks on Point Clouds]() `IEEE Internet of Things Journal`



## Defenses



- [The art of defense: letting networks fool the attacker.](https://arxiv.org/abs/2104.02963)  

  - [[Code](https://github.com/cuge1995/IT-Defense)]



- [DUP-Net: Denoiser and Upsampler Network for 3D Adversarial Point Clouds Defense.](https://arxiv.org/abs/1812.11017)  



- [IF-Defense: 3D Adversarial Point Cloud Defense via Implicit Function based Restoration.](https://arxiv.org/pdf/2010.05272)  

  - [[Code](https://github.com/Wuziyi616/IF-Defense)]



- [PointCutMix: Regularization Strategy for Point Cloud Classification.](https://arxiv.org/abs/2101.01461.pdf)  

  - [[Code](https://github.com/cuge1995/PointCutMix)]



- [PointGuard: Provably Robust 3D Point Cloud Classification.](https://arxiv.org/pdf/2103.03046.pdf) `CVPR 2021`



- [Defense-pointnet: Protecting pointnet against adversarial attacks.](https://arxiv.org/pdf/2002.11881)



- [LPF-Defense: 3D Adversarial Defense based on Frequency Analysis.](https://arxiv.org/pdf/2202.11287.pdf)  

  - [[Code](https://github.com/kimianoorbakhsh/LPF-Defence)]



- [Robust Structured Declarative Classifiers for 3D Point Clouds: Defending Adversarial Attacks with Implicit Gradients.](https://arxiv.org/pdf/2203.15245.pdf) `CVPR 2022`



- [Improving Adversarial Robustness of 3D Point Cloud Classification Models](https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136640663.pdf) `ECCV 2022`

  - [[Code](https://github.com/GuanlinLee/CCNAMS)]



- [Ada3Diff: Defending against 3D Adversarial Point Clouds via Adaptive Diffusion](https://arxiv.org/pdf/2211.16247.pdf)