Pusnet

This repo is the official code for

• Purified and Unified Steganographic Network

Dependencies and Installation

• Python 3.8.13, PyTorch = 1.11.0

• Run the following commands in your terminal:

  conda env create -f env.yml

  conda activate pyt_env

Get Started

Training

1. Change the code in config.py

   line4: mode = 'train'

   `line14: train_data_dir=''

   `line15: test_data_dir=''

2. Run python pusnet.py

• Trained models will be saved in 'model_zoo' folder.

Testing

1. Change the code in config.py

   line4: mode = 'test'

   line36-41: test_pusnet_path = ''

2. Run python pusnet.py

• Testing results will be saved in 'results' folder.
• Here, we provide trained models.

Dataset

• We train the PUSNet on the DIV2K training dataset, and test it on three testing datasets, including the DIV2K test dataset, 1000 images randomly selected from the ImageNet test dataset

Others

• The batch_size in config.py should be at least 2*number of gpus and it should be divisible by number of gpus.

Citation

If you find our paper or code useful for your research, please cite:

@inproceedings{li2024purified,
  title={Purified and Unified Steganographic Network},
  author={Li, Guobiao and Li, Sheng and Luo, Zicong and Qian, Zhenxing and Zhang, Xinpeng},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={27569--27578},
  year={2024}
}