The Pytorch version of Robust Defense Against Adversarial Steganography by Exploiting Differences Between CNN and Handcrafted Features.
CNN models in this repository, i.e. SiaStegNet and SID, are implemented in https://github.com/SiaStg/SiaStegNet.
Feature-based models and steganographic method come from http://dde.binghamton.edu/download/.
The Pytorch version of ADV-EMB was implemented by me. It is also published in https://github.com/ChuanQin/ADV-EMB-Pytorch.