Learning to Prevent Leakage: Privacy-Preserving Inference in the Mobile Cloud
This addresses privacy risks for users of mobile cloud applications by incrementally improving existing DNN-based systems.
The paper tackles the problem of privacy leakage when sending DNN features from mobile devices to the cloud by proposing a reinforcement learning framework that modifies DNN structures to prevent information leakage while maintaining high inference accuracy, with evaluations showing it successfully defends against various privacy attacks.
Powered by machine learning services in the cloud, numerous learning-driven mobile applications are gaining popularity in the market. As deep learning tasks are mostly computation-intensive, it has become a trend to process raw data on devices and send the deep neural network (DNN) features to the cloud, where the features are further processed to return final results. However, there is always unexpected leakage with the release of features, with which an adversary could infer a significant amount of information about the original data. We propose a privacy-preserving reinforcement learning framework on top of the mobile cloud infrastructure from the perspective of DNN structures. The framework aims to learn a policy to modify the base DNNs to prevent information leakage while maintaining high inference accuracy. The policy can also be readily transferred to large-size DNNs to speed up learning. Extensive evaluations on a variety of DNNs have shown that our framework can successfully find privacy-preserving DNN structures to defend different privacy attacks.