An efficient, ultrafast and flexible Python library for adversarial machine learning research, focusing on attacks, defenses, adversarial training, and benchmarking.
AdvRobust is a comprehensive Python library designed to facilitate research in adversarial machine learning. The library provides implementations of state-of-the-art adversarial attacks and defense mechanisms, with a particular emphasis on efficiency and flexibility for research experimentation.
This library is developed in conjunction with the following research paper:
Title: Adversarial examples attack based on random warm restart mechanism and improved Nesterov momentum
Authors: Tiangang Li
arXiv: https://arxiv.org/abs/2105.05029
BibTeX Citation:
@misc{li2021adversarialexamplesattackbased,
title={Adversarial examples attack based on random warm restart mechanism and improved Nesterov momentum},
author={Tiangang Li},
year={2021},
eprint={2105.05029},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2105.05029},
}- Implementation of various adversarial attack algorithms
- Defense mechanisms against adversarial examples
- Adversarial training frameworks
- Benchmarking tools for evaluating model robustness
- Efficient and optimized codebase for research experiments
pip install advrobust# Example usage coming soonWe welcome contributions to AdvRobust. Please refer to our contribution guidelines for more details.
This project is licensed under the MIT License - see the LICENSE file for details.