FANNG-MPC: Framework for Artificial Neural Networks and Generic MPC

Abstract

In this work, we introduce FANNG-MPC, a versatile secure multi-party computation framework capable to offer active security for privacy-preserving machine learning as a service (MLaaS). Derived from the now deprecated SCALE-MAMBA, FANNG is a data-oriented fork, featuring novel set of libraries and instructions for realizing private neural networks, effectively reviving the popular framework. To the best of our knowledge, FANNG is the first MPC framework to offer actively secure MLaaS in the dishonest majority setting. FANNG goes beyond SCALE-MAMBA by decoupling offline and online phases and materializing the dealer model in software, enabling a separate set of entities to produce offline material. The framework incorporates database support, a new instruction set for pre-processed material, including garbled circuits and convolutional and matrix multiplication triples. FANNG also implements novel private comparison protocols and an optimized library supporting Neural Network functionality. All our theoretical claims are substantiated by an extensive evaluation using an open-sourced implementation, including the private inference of popular neural networks like LeNet and VGG16.

Keywords

• Multi-Party Computation
• Privacy-Preserving Machine Learning
• Homomorphic Encryption
• Neural Networks
• MPC
• FHE