.png){kind=icon}
Return
Efficient privacy-preserving LSTM inference on encrypted sequential data
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2026.134508.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Recent advances in fully homomorphic encryption (FHE) have enabled privacy-preserving machine learning directly over encrypted data. As a representative recurrent architecture, the long short-term memory (LSTM) net work is widely used for modeling sequential dependencies, yet existing FHE-based LSTM inference schemes still suffer from high latency and limited scalability. In this paper, we present an efficient privacy-preserving LSTM in ference protocol on encrypted sequential data based on FHE. We insert a lightweight normalization module before nonlinear activations to bound the hidden states, thereby enabling accurate low-degree polynomial approxima tions of the Sigmoid, Tanh, and inverse-square-root functions via a hybrid Remez and least-squares strategy. We implement the proposed protocol using the Lattigo library, incorporating ciphertext packing optimization, rotation minimization, and SIMD-based parallelization. Experiments show that on standard text classification benchmarks, our encrypted LSTM achieves competitive accuracy compared with plaintext models and consistently outperforms the state-of-the-art FHE-based method, achieving up to a 4.7x speedup.
Keywords:
Machine learning as a service
Privacy preserving machine learning
Homomorphic encryption
LSTM
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W
Organization
Cited Papers
No cited papers available