One Noise to Rule Them All: Multi-View Adversarial Attacks with Universal Perturbation
Abstract: This paper presents a novel universal perturbation method for generating robust multi-view adversarial examples in 3D object recognition. Unlike conventional attacks limited to single views, our approach operates on multiple 2D images, offering a practical and scalable solution for enhancing model scalability and robustness. This generalizable method bridges the gap between 2D perturbations and 3D-like attack capabilities, making it suitable for real-world applications. Existing adversarial attacks may become ineffective when images undergo transformations like changes in lighting, camera position, or natural deformations. We address this challenge by crafting a single universal noise perturbation applicable to various object views. Experiments on diverse rendered 3D objects demonstrate the effectiveness of our approach. The universal perturbation successfully identified a single adversarial noise for each given set of 3D object renders from multiple poses and viewpoints. Compared to single-view attacks, our universal attacks lower classification confidence across multiple viewing angles, especially at low noise levels. A sample implementation is made available at https://github.com/memoatwit/UniversalPerturbation.
- Maksym Andriushchenko and Nicolas Flammarion. 2020. Understanding and improving fast adversarial training. Advances in Neural Information Processing Systems 33 (2020), 16048–16059.
- Frugal following: Power thrifty object detection and tracking for mobile augmented reality. In Proceedings of the 17th Conference on Embedded Networked Sensor Systems. 96–109.
- Alex Bes. 2017. Worn Baseball Ball. https://sketchfab.com/3d-models/worn-baseball-ball-fdf3de6ae225421ea78961b897b9608a Last accessed 10 February 2024.
- Nicholas Carlini and David Wagner. 2017. Towards Evaluating the Robustness of Neural Networks. arXiv:1608.04644Â [cs.CR]
- dannyboy70000. 2014. lemon 3D Model. https://free3d.com/3d-model/lemon-72357.html Last accessed 10 February 2024.
- Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition. Ieee, 248–255.
- Boosting Adversarial Attacks with Momentum. arXiv:1710.06081Â [cs.LG]
- Learning visual feature spaces for robotic manipulation with deep spatial autoencoders. arXiv preprint arXiv:1509.06113 25 (2015), 2.
- GetDeadEntertainment. 2020. Medieval Shovel. https://www.turbosquid.com/3d-models/medieval-shovel-3d-model-1494436 Last accessed 10 February 2024.
- Explaining and Harnessing Adversarial Examples. arXiv:1412.6572Â [stat.ML]
- Object recognition and robot grasping: A deep learning based approach. In The 34th Annual Conference of the Robotics Society of Japan (RSJ 2016), Yamagata, Japan.
- 3-D deformable object manipulation using deep neural networks. IEEE Robotics and Automation Letters 4, 4 (2019), 4255–4261.
- Roman Klokov and Victor Lempitsky. 2017. Escape from cells: Deep kd-networks for the recognition of 3d point cloud models. In Proceedings of the IEEE international conference on computer vision. 863–872.
- Adversarial examples in the physical world. arXiv:1607.02533Â [cs.CV]
- Truc Le and Ye Duan. 2018. Pointgrid: A deep network for 3d shape understanding. In Proceedings of the IEEE conference on computer vision and pattern recognition. 9204–9214.
- Stereo r-cnn based 3d object detection for autonomous driving. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 7644–7652.
- Object detection in the context of mobile augmented reality. In 2020 IEEE International Symposium on Mixed and Augmented Reality (ISMAR). IEEE, 156–163.
- Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks. arXiv:1908.06281Â [cs.LG]
- Edge assisted real-time object detection for mobile augmented reality. In The 25th annual international conference on mobile computing and networking. 1–16.
- Entangling metropolitan-distance separated quantum memories. arXiv preprint arXiv:2201.11953 (2022).
- Towards deep learning models resistant to adversarial attacks. arXiv preprint arXiv:1706.06083 (2017).
- mitsui. 2020. table 3D Model. https://free3d.com/3d-model/table-747735.html Last accessed 10 February 2024.
- Pointnet: Deep learning on point sets for 3d classification and segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition. 652–660.
- Multi-object detection and tracking, based on DNN, for autonomous vehicles: A review. IEEE Sensors Journal 21, 5 (2020), 5668–5677.
- Adversarial attacks and defenses in deep learning. Engineering 6, 3 (2020), 346–360.
- selfie 3DÂ scan. 2019. Tractor. https://sketchfab.com/3d-models/tractor-1b258bcc01bf4ed0935ef73e80442c30 Last accessed 10 February 2024.
- Intriguing properties of neural networks. arXiv preprint arXiv:1312.6199 (2013).
- NormalNet: A voxel-based CNN for 3D object classification and retrieval. Neurocomputing 323 (2019), 139–147.
- DNN based camera and LiDAR fusion framework for 3D object recognition. In Journal of Physics: Conference Series, Vol. 1518. IOP Publishing, 012044.
- Toward real-time 3D object recognition: A lightweight volumetric CNN framework using multitask learning. Computers & Graphics 71 (2018), 199–207.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.