-
Notifications
You must be signed in to change notification settings - Fork 1
/
Readme
21 lines (20 loc) · 1.51 KB
/
Readme
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
# Getting Started with DT-IBA
1. First you need to install the relevant installation packages, which are listed in requirements.txt, and run the following command from the command line to install them all.
pip install -r requirements.txt
2. Prepare pretrained model
Some parts we referenced excellent existing work, so you are invited to follow the work of the original authors if you need to.
The feature extraction phase of our identity hashing network uses a pre-trained SBS-R101, which can be obtained from FastReID's github repo (https://github.com/JDAI-CV/fast-reid/). In addition, the steganography network is referenced from StegaStamp and needs to be retrained according to the image size of the dataset (https://github.com/tancik/StegaStamp/).
@article{he2020fastreid,
title={FastReID: A Pytorch Toolbox for General Instance Re-identification},
author={He, Lingxiao and Liao, Xingyu and Liu, Wu and Liu, Xinchen and Cheng, Peng and Mei, Tao},
journal={arXiv preprint arXiv:2006.02631},
year={2020}
}
@inproceedings{2019stegastamp,
title={StegaStamp: Invisible Hyperlinks in Physical Photographs},
author={Tancik, Matthew and Mildenhall, Ben and Ng, Ren},
booktitle={IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2020}
}
3. Put the dataset to be poisoned into ". /dataset", and set the input and output path, parameters of the pre-trained model, and the configuration of the victim model in backdoor_implantation.py, then run the following command.
python backdoor_implantation.py