Efficient privacy-preserving scheme for secure neural network inference
Regular Papers|Updated:2025-10-13
|
Efficient privacy-preserving scheme for secure neural network inference
Enhanced Publication
用于安全神经网络推理的高效隐私保护方案
“In the field of cloud computing, this study introduces an efficient privacy-preserving secure neural network inference scheme. Expert xx proposed an approach of merging network parameters and a fast convolution algorithm, which reduces the linear operation time in the online stage by at least 11%, significantly reducing inference time and communication overhead.”
Frontiers of Information Technology & Electronic EngineeringVol. 26, Issue 9, Pages: 1609-1623(2025)
Affiliations:
1.School of Cyber Science and Engineering, Southeast University, Nanjing 210096, China
2.Purple Mountain Laboratories, Nanjing 211189, China