A generic framework for privacy preserving deep learning

2018/11/09 by Théo Ryffel, Theo Ryffel, Ryffel, Theo +13 · 1 voice · 342 citations
Computer Science · Mathematics · #Artificial intelligence #Computer science #Cryptography and Data Security #Deep learning #Internet Traffic Analysis and Secure E-voting #Internet privacy #Privacy-Preserving Technologies in Data #cs.CR #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1811.04017

published in arXiv (Cornell University) (Cornell University) · PPML 2018, 5 pages

openalex publication_date 2018/11/09 · arxiv created 2018/11/13 · arxiv updated 2018/11/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We detail a new framework for privacy preserving deep learning and discuss its assets. The framework puts a premium on ownership and secure processing of data and introduces a valuable representation based on chains of commands and tensors. This abstraction allows one to implement complex privacy preserving constructs such as Federated Learning, Secure Multiparty Computation, and Differential Privacy while still exposing a familiar deep learning API to the end-user. We report early results on the Boston Housing and Pima Indian Diabetes datasets. While the privacy features apart from Differential Privacy do not impact the prediction accuracy, the current implementation of the framework introduces a significant overhead in performance, which will be addressed at a later stage of the development. We believe this work is an important milestone introducing the first reliable, general framework for privacy preserving deep learning.

Cited by

• Security and Privacy Issues in Deep Learning
• Federated Learning in Mobile Edge Networks: A Comprehensive Survey
• FLASHE: Additively Symmetric Homomorphic Encryption for Cross-Silo Federated Learning
• Private Map-Secure Reduce: Infrastructure for Efficient AI Data Markets
• Federated Learning in Mobile Edge Networks: A Comprehensive Survey
• Flower: A Friendly Federated Learning Research Framework
• SoK: Training Machine Learning Models over Multiple Sources with Privacy Preservation
• Hierarchical Federated Learning through LAN-WAN Orchestration
• Partially Encrypted Machine Learning using Functional Encryption
• FedLab: A Flexible Federated Learning Framework
• Federated Learning With Differential Privacy: Algorithms and Performance Analysis
• Federated Learning for Big Data: A Survey on Opportunities, Applications, and Future Directions
• WAFFLE: Watermarking in Federated Learning
• FedBN: Federated Learning on Non-IID Features via Local Batch Normalization
• Prismo: A Decision Support System for Privacy-Preserving ML Framework Selection
• Robust Federated Learning by Mixture of Experts
• On the Impact of Device and Behavioral Heterogeneity in Federated Learning
• Differentially Private Bayesian Neural Networks on Accuracy, Privacy and Reliability
• SEALion: a Framework for Neural Network Inference on Encrypted Data
• Data Poisoning Attacks Against Federated Learning Systems
• SIRNN: A Math Library for Secure RNN Inference
• A Field Guide to Federated Optimization
• CrypTen: Secure Multi-Party Computation Meets Machine Learning
• FedCV: A Federated Learning Framework for Diverse Computer Vision Tasks
• CaRENets: Compact and Resource-Efficient CNN for Homomorphic Inference on Encrypted Medical Images
• A blockchain-orchestrated Federated Learning architecture for healthcare consortia
• PrivColl: Practical Privacy-Preserving Collaborative Machine Learning
• PowerTrip: Exploiting Federated Heterogeneous Datacenter Power for Distributed ML Training
• Characterizing Linear Alignment Across Language Models
• Asymmetric Private Set Intersection with Applications to Contact Tracing and Private Vertical Federated Machine Learning
• Advances and Open Problems in Federated Learning
• Fed-Focal Loss for imbalanced data classification in Federated Learning
• Privacy-Preserving Distributed Learning Framework for 6G Telecom Ecosystems
• IBM Federated Learning: an Enterprise Framework White Paper V0.1
• AdeptHEQ-FL: Adaptive Homomorphic Encryption for Federated Learning of Hybrid Classical-Quantum Models with Dynamic Layer Sparing
• TiFL: A Tier-based Federated Learning System
• Practical Defences Against Model Inversion Attacks for Split Neural Networks
• Privacy-preserving medical image analysis
• GFL: A Decentralized Federated Learning Framework Based On Blockchain
• A Federated Random Forest Solution for Secure Distributed Machine Learning
• Characterizing Impacts of Heterogeneity in Federated Learning upon Large-Scale Smartphone Data
• Reliability and Performance Assessment of Federated Learning on Clinical Benchmark Data
• A Framework for Democratizing AI
• ByzFL: Research Framework for Robust Federated Learning
• FedJAX: Federated learning simulation with JAX
• Privacy-preserving Decentralized Aggregation for Federated Learning
• Federated Learning for Healthcare Informatics
• PyVertical: A Vertical Federated Learning Framework for Multi-headed SplitNN
• Technology readiness levels for machine learning systems
• FedEval: A Holistic Evaluation Framework for Federated Learning
• Real-World Image Datasets for Federated Learning
• Federated Learning System without Model Sharing through Integration of Dimensional Reduced Data Representations
• Towards Communication-Efficient and Attack-Resistant Federated Edge Learning for Industrial Internet of Things
• Self-supervised Cross-silo Federated Neural Architecture Search
• Fast-adapting and Privacy-preserving Federated Recommender System

Discussions

• Paper: A Generic Framework for Privacy Preserving Deep Learning (NIPS 2018) [hn, 25 points, 1 comments]

Related

• Security and Privacy Preserving Deep Learning
• Efficient Deep Learning on Multi-Source Private Data
• Recent Advances of Differential Privacy in Centralized Deep Learning: A Systematic Survey
• Improving the Algorithm of Deep Learning with Differential Privacy
• An Adaptive and Fast Convergent Approach to Differentially Private Deep Learning