Code for the paper "A4NT: Author Attribute Anonymity by Adversarial Training of Neural Machine Translation" presented in Usenix 2018. Link
Pre-processed datasets used in our paper are available below in json format
Pre-trained translator models are available below:
- Age translator for adult (23-40 years) to teenager(13-19 years) and vice-versa.
- Gender translator
- Speech translator for Obama to Trump and vice-versa.
Code is written in python 2.7 and uses pytorch 3.x library.
More instructions coming soon
If you find this code useful in your work, please cite the paper.
@inproceedings{shetty_USENIX2018,
TITLE = {$A^{4}NT$: Author Attribute Anonymity by Adversarial Training of Neural Machine Translation},
AUTHOR = {Shetty, Rakshith and Schiele, Bernt and Fritz, Mario},
LANGUAGE = {eng},
ISBN = {978 -1- 931971- 4 6 -1},
PUBLISHER = {USENIX Association},
YEAR = {2018},
BOOKTITLE = {Proceedings of the 27th USENIX Security Symposium},
PAGES = {1633--1650},
ADDRESS = {Baltimore, MD, USA},
}