Paper 2021/1100

REDsec: Running Encrypted Discretized Neural Networks in Seconds

Lars Folkerts, Charles Gouert, and Nektarios Georgios Tsoutsos


Machine learning as a service (MLaaS) has risen to become a prominent technology due to the large development time, amount of data, hardware costs, and level of expertise required to develop a machine learning model. However, privacy concerns prevent the adoption of MLaaS for applications with sensitive data. A promising privacy preserving solution is to use fully homomorphic encryption (FHE) to perform the ML computations. Recent advancements have lowered computational costs by several orders of magnitude, opening doors for secure practical applications to be developed. This work looks to optimize FHE-based private machine learning inference by leveraging ternary neural networks. Such neural networks, whose weights are constrained to {-1,0,1}, have special properties that we exploit in this work to operate efficiently in the homomorphic domain. We introduce a general framework that takes a user-defined model as input (bring-your-own-network), performs plaintext training, and efficiently evaluates private inference leveraging FHE. We perform inference experiments with the MNIST, CIFAR-10, and ImageNet datasets and achieve speeds as fast as 1.6 to 2.2 orders of magnitude slower than unencrypted single-core performance.

Available format(s)
Publication info
Preprint. Minor revision.
fully homomorphic encryptionGPU accelerationencrypted neural networksprivacy-preserving inference
Contact author(s)
tsoutsos @ udel edu
2022-04-29: revised
2021-08-26: received
See all versions
Short URL
Creative Commons Attribution


      author = {Lars Folkerts and Charles Gouert and Nektarios Georgios Tsoutsos},
      title = {REDsec: Running Encrypted Discretized Neural Networks in Seconds},
      howpublished = {Cryptology ePrint Archive, Paper 2021/1100},
      year = {2021},
      note = {\url{}},
      url = {}
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.