Large-capacity and Flexible Video Steganography via Invertible Neural Network (CVPR 2023)

Chong Mou, Youmin Xu, Jiechong Song, Chen Zhao, Bernard Ghanem, Jian Zhang

Official implementation of Large-capacity and Flexible Video Steganography via Invertible Neural Network.

Introduction

Video steganography is the art of unobtrusively concealing secret data in a cover video and then recovering the secret data through a decoding protocol at the receiver end. Although several attempts have been made, most of them are limited to low-capacity and fixed steganography. To rectify these weaknesses, we propose a Large-capacity and Flexible Video Steganography Network (LF-VSN) in this paper. For large-capacity, we present a reversible pipeline to perform multiple videos hiding and recovering through a single invertible neural network (INN). Our method can hide/recover 7 secret videos in/from 1 cover video with promising performance. For flexibility, we propose a key-controllable scheme, enabling different receivers to recover particular secret videos from the same cover video through specific keys. Moreover, we further improve the flexibility by proposing a scalable strategy in multiple videos hiding, which can hide variable numbers of secret videos in a cover video with a single model and a single training session. Extensive experiments demonstrate that with the significant improvement of the video steganography performance, our proposed LF-VSN has high security, large hiding capacity, and flexibility.

---

🔧 Dependencies and Installation

• Python 3.6
• PyTorch >= 1.4.0
• numpy
• skimage
• cv2

⏬ Download Models

The pre-trained models are available at:

Mode	Download link
One video hiding	Google Drive
Two video hiding	Google Drive
Three video hiding	Google Drive
Four video hiding	Google Drive
Five video hiding	Google Drive
Six video hiding	Google Drive
Seven video hiding	Google Drive

Data Preparing

Please download the training and evaluation dataset from Vimeo-90K.

Train

Training the desired model by changing the config file.

python train.py -opt options/train/train_LF-VSN_1video.yml

Test

Testing the desired model by changing the config file.

python test.py -opt options/train/train_LF-VSN_1video.yml

Qualitative Results

🤗 Acknowledgements

This code is built on MIM-VRN (PyTorch). We thank the authors for sharing their codes of MIMO-VRN.

:e-mail: Contact

If you have any question, please email eechongm@gmail.com.

Citation

If you find our work helpful in your resarch or work, please cite the following paper.

@inproceedings{mou2023lfvsn,
  title={Large-capacity and Flexible Video Steganography via Invertible Neural Network},
  author={Chong Mou, Youmin Xu, Jiechong Song, Chen Zhao, Bernard Ghanem, Jian Zhang},
  booktitle={CVPR},
  year={2023}
}