Deep Quaternion Features for Privacy Protection
This addresses privacy protection for users of neural networks, particularly in scenarios where intermediate features might be exposed, though it appears incremental as it builds on existing neural network architectures with a quaternion-based twist.
The paper tackles the problem of preventing input information leakage from intermediate-layer features in neural networks by proposing a quaternion-valued neural network (QNN) that hides input data in a random phase, achieving effective privacy protection with only mild accuracy degradation and lower computational cost compared to other methods.
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.