Announcing the OpenMined-UCSF Data-Centric Federated Learning Fellowship

We’re very excited to announce the next round of open-source software development grants in the OpenMined community, generously sponsored by the University of California San Francisco! These grants will focus on bringing data-centric federated learning with differential privacy budgeting to PyGrid.

What is Secure Multi-Party Computation?

This post is part of our Privacy-Preserving Data Science, Explained Simply series.

Privacy-Preserving Data Science, Explained

In this blog series, we’ll explain common topics in privacy-preserving data science, from a single sentence to code examples. We hope these posts serve as a useful resource for you to figure out the best techniques for your organization.

What is Federated Learning?

This post is part of our Privacy-Preserving Data Science, Explained series. In this article of the introductory series on Private ML, we wi

Privacy-Preserving AI Summary: MIT Deep Learning Series

This article briefly discusses the key concepts covered in Part 1 of this lecture by Andrew Trask, which is part of the MIT Deep Learning Se

PySyft, PyTorch and Intel SGX: Secure Aggregation on Trusted Execution Environments

When we talk about sensitive data and cloud computing, how can we guarantee the remote, secure and private execution of our applications? One possible solution to this problem is to run the application in a Trusted Execution Environment (TEE).

PyGrid: A Peer-to-Peer Platform for Private Data Science and Federated Learning

What if you could train on all of the world’s data, without that data leaving the device, and while keeping that data private? PyGrid is a peer-to-peer platform for private data science and federated learning.

Inference privacy: what is it, and why do we care?

When you ask your home voice assistant to check the weather, it is listening to and saving not only your voice but also everything else tha

Using Privacy and Federated Learning in Recommendations - Part 1

Recommendation systems are everywhere in our everyday life online — they can be incredibly useful, time-saving, and aid in our discovery of things relevant to our interests. Privacy-preserving recommendation systems can use better signals to build better models.

Autonomous Driving's Seat Belt Moment

Volvo made seatbelt patents readily available to all competitors to encourage adoption of this life saving innovation. Is there a way to promote sharing, or discovery, of safety critical innovations in autonomous vehicles?

Predictive Maintenance of Turbofan Engines using Federated Learning with PySyft and PyGrid

Is it possible to benefit from the wonders of machine learning without having direct access to data? Today, machine learning can be used to

Introduction to Federated Learning and Privacy Preservation using PySyft and PyTorch

Federated Learning and Additive Secret Sharing using the PySyft framework

Split Neural Networks on PySyft and PyTorch

Summary: In this blog we are going to provide an introduction into a new decentralised learning methodology called, ‘Split Neural Networks’.

Meet OpenMined's new PyTorch-OpenMined Fellows

We’re very excited to announce the recipients of the latest round of open-source software development grants in the OpenMined community, gen

Announcing the OpenMined-PyTorch Federated Learning Fellowships

We’re very excited to announce the next round of grants sponsored by the PyTorch team! This grant will focus on developing “worker libraries”, allowing PySyft code to be executed in other environments like a mobile phone or web browser.

Introducing PySyft TensorFlow

Support for TensorFlow in PySyft!

Asynchronous Federated Learning in PySyft

In this post, we provide a showcase of applying federated learning using PySyft. PySyft is an open source python library for secure and priv

Privacy-Preserving AI in Medical Imaging: Federated Learning, Differential Privacy, and Encrypted Computation

In medical imaging, necessary privacy concerns limit us from fully maximizing the benefits of AI in our research. These modern privacy techniques could allow us to train our models on encrypted data from multiple institutions, hospitals, and clinics without sharing the patient data.

Encrypted Training with PyTorch + PySyft

Encrypted Training of Deep Learning models with PyTorch + PySyft on MNIST

Federated Learning of a Recurrent Neural Network on Raspberry PIs

In this article, you are going to learn how to setup PySyft on a Raspberry PI and how to train a Recurrent Neural Network in a federated way.

Encrypted Deep Learning Classification with PyTorch & PySyft

Encrypted Deep Learning Classification with PyTorch & PySyft in < 33ms on MNIST

Deep Learning -> Federated Learning in 10 Lines of PyTorch + PySyft

Deep Learning -> Federated Learning in 10 Lines of PyTorch + PySyft

Weekly Digs #10

Small but good: we only dug up one paper this week but it comes with very interesting claims.

Weekly Digs #9

If anyone had any doubt that private machine learning is a growing area then this might take care of that: one week with papers on MPC, HE, SGX, and DP.