Personal photos of individuals when shared online, apart from exhibiting a myriad of memorable details, also reveals a wide range of private information and potentially entails privacy risks (e.g., online harassment, tracking). To mitigate such risks, it is crucial to study techniques that allow individuals to limit the private information leaked in visual data. We tackle this problem in a novel image obfuscation framework: to maximize entropy on inferences over targeted privacy attributes, while retaining image fidelity. We approach the problem based on an encoder-decoder style architecture, with two key novelties: (a) introducing a discriminator to perform bi-directional translation simultaneously from multiple unpaired domains; (b) predicting an image interpolation which maximizes uncertainty over a target set of attributes. We find our approach generates obfuscated images faithful to the original input images, and additionally increase uncertainty by 6.2× (or up to 0.85 bits) over the non-obfuscated counterparts.
@article{arxiv.2005.10329,
title = {InfoScrub: Towards Attribute Privacy by Targeted Obfuscation},
author = {Hui-Po Wang and Tribhuvanesh Orekondy and Mario Fritz},
journal= {arXiv preprint arXiv:2005.10329},
year = {2021}
}