Paper 2020/050

Delphi: A Cryptographic Inference Service for Neural Networks

Pratyush Mishra, Ryan Lehmkuhl, Akshayaram Srinivasan, Wenting Zheng, and Raluca Ada Popa

Abstract

Many companies provide neural network prediction services to users for a wide range of applications. However, current prediction systems compromise one party's privacy: either the user has to send sensitive inputs to the service provider for classification, or the service provider must store its proprietary neural networks on the user's device. The former harms the personal privacy of the user, while the latter reveals the service provider's proprietary model. We design, implement, and evaluate Delphi, a secure prediction system that allows two parties to execute neural network inference without revealing either party's data. Delphi approaches the problem by simultaneously co-designing cryptography and machine learning. We first design a hybrid cryptographic protocol that improves upon the communication and computation costs over prior work. Second, we develop a planner that automatically generates neural network architecture configurations that navigate the performance-accuracy trade-offs of our hybrid protocol. Together, these techniques allow us to achieve a 22x improvement in online prediction latency compared to the state-of-the-art prior work.

Metadata
Available format(s)
PDF
Category
Cryptographic protocols
Publication info
Published elsewhere. Minor revision. USENIX Security 2020
Keywords
privacy-preserving machine learningdeep learningsecure inferenceneural architecture search
Contact author(s)
pratyush @ berkeley edu
raluca popa @ berkeley edu
History
2020-05-07: revised
2020-01-17: received
See all versions
Short URL
https://ia.cr/2020/050
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2020/050,
      author = {Pratyush Mishra and Ryan Lehmkuhl and Akshayaram Srinivasan and Wenting Zheng and Raluca Ada Popa},
      title = {Delphi: A Cryptographic Inference Service for Neural Networks},
      howpublished = {Cryptology ePrint Archive, Paper 2020/050},
      year = {2020},
      note = {\url{https://eprint.iacr.org/2020/050}},
      url = {https://eprint.iacr.org/2020/050}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.