PrivLSTM: A Privacy-Preserving LSTM Inference Framework by Fusing Encryption and Network Structure for Multi-Sourced Data

doi:10.1016/j.inffus.2025.103711

PrivLSTM: A Privacy-Preserving LSTM Inference Framework by Fusing Encryption and Network Structure for Multi-Sourced Data

2025-09-11

0

PRE

AI

DOI：10.1016/j.inffus.2025.103711

Original

Original request for help

Save

Abstract

En 中文

• PrivLSTM protects the user’s private input sequences while being user-friendly. • A batch-based linear transformation method to reduce the automorphism operation. • An optimized approach to approximate the nonlinear activation function. • An integration operation to improve the computation efficiency.

PrivLSTM: A Privacy-Preserving LSTM Inference Framework by Fusing Encryption and Network Structure for Multi-Sourced Data

Abstract

Journal

Organization