CrypTFlow: Secure TensorFlow Inference

By Nishant Kumar, Mayank Rathee, Nishanth Chandran, Divya Gupta, Aseem Rastogi, Rahul Sharma. 2019

280 citations recorded by the source corpus.

We present CrypTFlow, a first of its kind system that converts TensorFlow inference code into Secure Multi-party Computation (MPC) protocols at the push of a button. To do this, we build three components. Our first component, Athos, is an end-to-end compiler from TensorFlow to a variety of semihonest MPC protocols. The second component, Porthos, is an improved semi-honest 3-party protocol that provides significant speedups for TensorFlow like applications. Finally, to provide malicious secure MPC protocols, our third component, Aramis, is a novel technique that uses hardware with integrity guarantees to convert any semi-honest MPC protocol into an MPC protocol that provides malicious security. The malicious security of the protocols output by Aramis relies on integrity of the hardware and semi-honest security of MPC. Moreover, our system matches the inference accuracy of plaintext TensorFlow.We experimentally demonstrate the power of our system by showing the secure inference of real-world neural networks such as ResNet50 and DenseNet121 over the ImageNet dataset with running times of about 30 seconds for semi-honest security and under two minutes for malicious security. Prior work in the area of secure inference has been limited to semi-honest security of small networks over tiny datasets such as MNIST or CIFAR. Even on MNIST/CIFAR, CrypTFlow outperforms prior work.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Privacy in Deep Learning: A Survey (2020, neutral): “ARDEN [128] # # # MNIST, CIFAR-10, SVHN Image Classification w/ DNN Cryptonets [129] # # # MNIST Image Classification w/ DNN Private Classification [130] # # # MNIST Image Classification w/ DNN TAPAS [131] # # # MNIST, Faces, Cancer, DiabetesImage Classification w/ DNN FHE–DiNN [132] # # # MNIST Ima”
Secure Evaluation of Quantized Neural Networks (2019, methodology): “Moreover, the CrypTFlow framework [44] also support secure inference with malicious security albeit by relying on a secure hardware assumption.”
MP-SPDZ: A Versatile Framework for Multi-Party Computation (2020, methodology): “These are compiled automatically from TensorFlow with the help of CrypTFlow [56], which opens the possibility for further networks to be used.”
Cheetah: Lean and Fast Secure Two-Party Deep Neural Network Inference (2022, background): “For instance, [40] is more than 15× faster than CrypTFlow2 on ResNet50.”
CrypTFlow2: Practical 2-Party Secure Inference (2020, methodology): “Equivalently, ReLU ( 𝑎 ) = DReLU ( 𝑎 ) · 𝑎 .”
Trustworthy AI: From Principles to Practices (2021, methodology): “Several tools have been open sourced, such as MP2ML [48], CryptoSPN [330], CrypTFlow [200, 276], and CrypTen [188].”
Secure, privacy-preserving and federated machine learning in medical imaging (2020, background): “…and scalability to more than a small number of parties is a concern for SMPC applications 77 , and computational considerations are a concern beyond small algorithm sizes, with efficient SMPC implementations of state-of-the-art neural network algorithms currently under active development 78 .”
CrypTen: Secure Multi-Party Computation Meets Machine Learning (2021, methodology): “Other systems that work in this setting include Astra [16], Blaze [56], and CrypTFlow [41].”
Falcon: Honest-Majority Maliciously Secure Framework for Private Deep Learning (2020, unknown): “CryptFlow [13] # # # # G # # # G#”