InfoScrub: Towards Attribute Privacy by Targeted Obfuscation
arXiv:2005.10329
Abstract
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.
arXiv comment20 pages, 7 figures
References in corpus (8)
- Adam: A Method for Stochastic Optimization
- Conditional Generative Adversarial Nets
- mixup: Beyond Empirical Risk Minimization
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Spectral Normalization for Generative Adversarial Networks
- Conditional Image Synthesis With Auxiliary Classifier GANs
- Invertible Conditional GANs for image editing
- Flexibly Fair Representation Learning by Disentanglement