Summary of Lidattack: Robust Black-box Attack on Lidar-based Object Detection, by Jinyin Chen et al.
LiDAttack: Robust Black-box Attack on LiDAR-based Object Detection
by Jinyin Chen, Danxin Liao, Sheng Xiang, Haibin Zheng
First submitted to arxiv on: 4 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research proposes a robust black-box attack called LiDAttack that targets LiDAR sensors used in object detection models. The attack uses a genetic algorithm with simulated annealing to limit the location and number of perturbation points, making it stealthy and effective. It also simulates scanning deviations to adapt to real-world scenario variations. Experiments are conducted on three datasets (KITTI, nuScenes, and self-constructed data) using three dominant object detection models (PointRCNN, PointPillar, and PV-RCNN++). The results show the attack’s efficiency in targeting a wide range of object detection models, with an attack success rate (ASR) up to 90%. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper introduces a new type of attack that can be used on LiDAR sensors. It uses a special algorithm to make the attack work well and is able to adapt to changes in the real world. The researchers tested their attack on different datasets and object detection models, and it was very successful. |
Keywords
» Artificial intelligence » Object detection » Rcnn
