Training data is often stored on a single machine or in a data centre in traditional machine learning approaches. Federated learning is a privacy-preserving technique that comes in handy when the training data is scarce, confidential, or lacking in variety.
NVIDIA FLARE, which stands for Federated Learning Application Runtime Environment, is an open-source project by NVIDIA. It’s a software development kit that allows people from all around the world to work together to create more generalizable AI models. NVIDIA FLARE is the fundamental engine in NVIDIA Clara Train’s federated learning software, which has been used in medical imaging, genetic analysis, cancer research, and COVID-19 research.
Researchers can modify their method for domain-specific applications by choosing from a variety of federated learning architectures. NVIDIA FLARE can also be used by platform developers to give consumers with the distributed infrastructure needed to create a multi-party collaborative application.
Two federated learning projects have employed NVIDIA FLARE:
NVIDIA and Roche Digital Pathology researchers cooperated on a successful internal simulation that classified full slide images.
It also collaborated on an AI application with Erasmus Medical Center in the Netherlands to detect genetic variations linked to schizophrenia patients.
The server-client architecture, on the other hand, is not suitable for all federated learning applications. By supporting additional architectures, NVIDIA FLARE will make federated learning more accessible to a wider range of applications. Among the conceivable applications are:
- Seismic and wellbore data are analyzed by energy companies.
- Financial institutions are working to improve their fraud detection algorithms.
- Manufacturers optimizing industrial processes
Current AI initiatives, such as the open-source MONAI medical imaging platform, can interact with NVIDIA FLARE. It will also be used to power federated learning solutions in the following areas:
- The American College of Radiography (ACR) and NVIDIA cooperated on federated learning research for breast cancer and COVID-19 applications using artificial intelligence. It intends to make NVIDIA FLARE available through the ACR AI-LAB, a software platform that tens of thousands of society members use.
- Users can use Flywheel Exchange to access and exchange data and biomedical research techniques, administer federated projects for training and research, and choose their preferred federated learning system, such as NVIDIA FLARE.
- Rhino Health, an NVIDIA Inception programme partner, has integrated NVIDIA FLARE into their federated learning system, aiding Massachusetts General Hospital researchers in the development of an AI model that more accurately diagnoses brain aneurysms. It also aids the National Cancer Institute’s Early Detection Research Network in the development and validation of medical imaging AI models that detect early symptoms of pancreatic cancer.