mlsanityData to 7 Oct 2026

Glyph: Fast and Accurately Training Deep Neural Networks on Encrypted Data

arXiv 1911.07101 ↗First posted 16 Nov 2019cs.LG stat.MLcited by 28

Not on the map. Why?

Abstract (arXiv)

Big data is one of the cornerstones to enabling and training deep neural networks (DNNs). Because of the lack of expertise, to gain benefits from their data, average users have to rely on and upload their private data to big data companies they may not trust. Due to the compliance, legal, or privacy constraints, most users are willing to contribute only their encrypted data, and lack interests or resources to join the training of DNNs in cloud. To train a DNN on encrypted data in a completely non-interactive way, a recent work proposes a fully homomorphic encryption (FHE)-based technique implementing all activations in the neural network by textitBrakerski-Gentry-Vaikuntanathan (BGV)-based lookup tables. However, such inefficient lookup-table-based activations significantly prolong the training latency of privacy-preserving DNNs. In this paper, we propose, Glyph, a FHE-based scheme to fast and accurately train DNNs on encrypted data by switching between TFHE (Fast Fully Homomorphic Encryption over the Torus) and BGV cryptosystems. Glyph uses logic-operation-friendly TFHE to implement nonlinear activations, while adopts vectorial-arithmetic-friendly BGV to perform multiply-accumulation (MAC) operations. Glyph further applies transfer learning on the training of DNNs to improve the test accuracy and reduce the number of MAC operations between ciphertext and ciphertext in convolutional layers. Our experimental results show Glyph obtains the state-of-the-art test accuracy, but reduces the training latency by 99% over the prior FHE-based technique on various encrypted datasets.

Key facts. Cited by 28 papers in the map (data through 7 Oct 2026). First posted on arXiv 16 Nov 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 6; 2022: 2; 2023: 3; 2024: 6; 2025: 8; 2026: 3.

28 papers.

Sort:

By area:

  1. 2022Trustworthy Graph Neural Networks: Aspects, Methods and Trendscited by 88
  2. 2021Differential Privacy for Text Analytics via Natural Text Sanitizationcited by 74
  3. 2021Privacy-Preserving Machine Learning: Methods, Challenges and Directionscited by 58
  4. 2021SIRNN: A Math Library for Secure RNN Inferencecited by 28
  5. 2021Secure Quantized Training for Deep Learningcited by 15
  6. 2021A methodology for training homomorphicencryption friendly neural networkscited by 10
  7. 2025CipherPrune: Efficient and Scalable Private Transformer Inferencecited by 9
  8. 2024GuardML: Efficient Privacy-Preserving Machine Learning Services Through Hybrid Homomorphic Encryptioncited by 7
  9. 2025Private LoRA Fine-tuning of Open-Source LLMs with Homomorphic Encryptioncited by 5
  10. 2021Adam in Private: Secure and Fast Training of Deep Neural Networks with Adaptive Moment Estimationcited by 4
  11. 2025TFHE-Coder: Evaluating LLM-agentic Fully Homomorphic Encryption Code Generationcited by 4
  12. 2024CryptoTrain: Fast Secure Training on Encrypted Datasetcited by 3
16 more
  1. 2025RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data Synthesiscited by 3
  2. 2024MedBlindTuner: Towards Privacy-preserving Fine-tuning on Biomedical Images with Transformers and Fully Homomorphic Encryptioncited by 2
  3. 2024Neural Network Training on Encrypted Data with TFHEcited by 2
  4. 2025ReBoot: Encrypted Training of Deep Neural Networks with CKKS Bootstrappingcited by 1
  5. 2025SoK: Data Minimization in Machine Learningcited by 1
  6. 2025DictPFL: Efficient and Private Federated Learning on Encrypted Gradientscited by 1
  7. 2022Protecting Data from all Parties: Combining FHE and DP in Federated Learningcited by 0
  8. 2023Data Privacy with Homomorphic Encryption in Neural Networks Training and Inferencecited by 0
  9. 2023SABLE: Secure And Byzantine robust LEarningcited by 0
  10. 2023Practical, Private Assurance of the Value of Collaboration via Fully Homomorphic Encryptioncited by 0
  11. 2024I can't see it but I can Fine-tune it: On Encrypted Fine-tuning of Transformers using Fully Homomorphic Encryptioncited by 0
  12. 2024Homomorphic WiSARDs: Efficient Weightless Neural Network training over encrypted datacited by 0
  13. 2025Functional Encryption in Secure Neural Network Training: Data Leakage and Practical Mitigationscited by 0
  14. 2026Continual Pretraining on Encrypted Synthetic Data for Privacy-Preserving LLMscited by 0
  15. 2026Detecting and Understanding Vulnerabilities in Fully Homomorphic Encryption Frameworkscited by 0
  16. 2026Beyond Classification: Task-Dependent Learnability under Privacy-Motivated Image Transformationscited by 0