Hide in Thicket: Generating Imperceptible and Rational Adversarial Perturbations on 3D Point Clouds (CVPR 2024)
This work corresponds to the following paper: https://arxiv.org/abs/2403.05247.
This repo is based on [https://github.com/code-roamer/AOF.git] and [https://github.com/shikiw/SI-Adv.git] to perform adversarial attack.
@article{lou2024HiT-ADV,
title={Hide in Thicket: Generating Imperceptible and Rational Adversarial Perturbations on 3D Point Clouds},
author={Lou, Tianrui and Jia, Xiaojun and Gu, Jindong and Liu, Li and Liang, Siyuan and He, Bangyan and Cao, Xiaochun},
journal={arXiv preprint arXiv:2403.05247},
year={2024}
}
Step 1. Create a conda environment or use your existing one.
conda create --name hitadv python=3.8 -y
conda activate hitadv
pip install -r requirements.txt
Step 2. Prepare datasets and pretrained models. Download from Baidu Yun：https://pan.baidu.com/s/1SL5-TuT9n74x5mADSM2E9g (password:eaic)
Step 3. Evaluating：
python eval.py
Visualizing:
python visual.py