This is an PyTorch implementation of AAAI 2021 paper DeHiB: Deep Hidden Backdoor Attack on Semi-supervised Learning via Adversarial Perturbation.
This code is only available in FixMatch (RandAugment). Now only experiments on CIFAR-10 and CIFAR-100 are available.
- Python
- PyTorch
- torchvision
- tqdm
- numpy
- yattag
- pandas
- sklearn
- matplotlib
- Pillow
If you find this code is useful for your research, please cite our paper:
@article{yan2021dehib,
title={DeHiB: Deep Hidden Backdoor Attack on Semi-supervised Learning via Adversarial Perturbation},
journal={Proceedings of the AAAI Conference on Artificial Intelligence},
author={Yan, Zhicong and Li, Gaolei and TIan, Yuan and Wu, Jun and Li, Shenghong and Chen, Mingzhe and Poor, H. Vincent},
year={2021},
pages={10585-10593}
}
If you have any questions, drop an email to zhicongy@sjtu.edu.cn