Cs-FNNS

This repo is the official code for

• Cover-separable Fixed Neural Network Steganography via Deep Generative Models

Dependencies and Installation

• Python 3.8.13, PyTorch = 1.11.0

• Run the following commands in your terminal:

  conda env create -f env.yaml

  conda activate FNNS

  pip install natsort

Get Started

• Regarding resistance against detection,

  Run python Cs-FNNS_AntiDetect.py

• Regarding resistance against JPEG compression,

  Change the code in config.py: line14: secret_image_size = '128'

  Run python Cs-FNNS-JPEG.py

• Regarding hiding multiple secret images for different receiver,

  Run python Cs-FNNS_MUsers.py

• Results will be saved in the "./result" folder.

Others

• Pretrained SRNet model is large, so we upload it to Google Drive. Download it and place it in the specified location (refer to line 20 in the config.py file). After that, you can uncomment the part related to SRNet, bringing its gradient signals into the SPS optimization.

Citation

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

@inproceedings{li2024cover,
  title={Cover-separable Fixed Neural Network Steganography via Deep Generative Models},
  author={Li, Guobiao and Li, Sheng and Qian, Zhenxing and Zhang, Xinpeng},
  booktitle={Proceedings of the 32nd ACM International Conference on Multimedia},
  pages={10238--10247},
  year={2024}
}