OpenFL (Federated Learning) Building better AI models with private data
summary
OpenFL is an open source federated learning framework that runs an aggregator-as-certificate-authority over gRPC and collaborators that train local models and send clipped weight deltas for aggregation; the demo uses the FX CLI, a plan.yaml with hyperparameters, sharded MNIST data and PyTorch model, and mutual TLS via built‑in PKI. Security add‑ons use Intel SGX plus remote attestation and global differential privacy; published studies train on MRI scans from 6,000+ patients across seven sites and report a 33% improvement in tumor detection.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:20Why Federated Learning
- 3:05OpenFL Architecture
- 5:00Demo
- 9:05Certificate Authority
- 14:35Security and Privacy
- 15:55OpenFL Security
- 17:25GitHub
- 17:58Data Diversity
- 20:17Getting started
- 21:09Questions
related talks
39mOpenFL: Building Better AI Models with Private Data
32mOpenFL: A Federated Learning Project to Power (and Secure) Your Projects
31mFrom Theory to Practice: Advances in Real-World Federated Learning
30mFederated Learning from simulation to production
42mFederated Learning: Unlocking the Values of Discrete Data for Enterprise
31mFederated Learning: So Close and Yet So Out of Reach
27mCreate Value and Insight Through Privacy-preserving Data Collaboration [...]
18mEnabling Medical AI through Federated Learning