Deep Quaternion Features for Privacy Protection

Abstract

We propose a method to revise the neural network to construct the quaternion-valued neural network (QNN), in order to prevent intermediate-layer features from leaking input information. The QNN uses quaternion-valued features, where each element is a quaternion. The QNN hides input information into a random phase of quaternion-valued features. Even if attackers have obtained network parameters and intermediate-layer features, they cannot extract input information without knowing the target phase. In this way, the QNN can effectively protect the input privacy. Besides, the output accuracy of QNNs only degrades mildly compared to traditional neural networks, and the computational cost is much less than other privacy-preserving methods.

References in corpus (8)

• Deep Complex Networks
• Associative Long Short-Term Memory
• Quaternion Recurrent Neural Networks
• PrivyNet: A Flexible Framework for Privacy-Preserving Deep Neural Network Training
• Not Just Privacy: Improving Performance of Private Deep Learning in Mobile Cloud
• Adversarial Neural Network Inversion via Auxiliary Knowledge Alignment
• Interpretable Complex-Valued Neural Networks for Privacy Protection
• Rotation-Equivariant Neural Networks for Privacy Protection

Cited by in corpus (1)

• Rotation-Equivariant Neural Networks for Privacy Protection