You Can Use But Cannot Recognize: Preserving Visual Privacy in Deep Neural Networks

arXiv:2404.04098v1 Announce Type: new
Abstract: Image data have been extensively used in Deep Neural Network (DNN) tasks in various scenarios, e.g., autonomous driving and medical image analysis, which incurs significant privacy concerns. Existing privacy protection techniques are unable to efficiently protect such data. For example, Differential Privacy (DP) that is an emerging technique protects data with strong privacy guarantee cannot effectively protect visual features of exposed image dataset. In this paper, we propose a novel privacy-preserving framework VisualMixer that protects …

analysis arxiv autonomous autonomous driving can cs.cr data differential privacy driving image image analysis medical network networks neural network neural networks privacy privacy concerns protect protection techniques