Research Papers

NVIDIA FLARE has been used in peer-reviewed research across medical imaging, federated optimization, privacy analysis, and large language models. Reference implementations are available in the research directory on GitHub.

Published Research

• FedNCA – Equitable Federated Learning with NCA (MICCAI 2025)

• FedBPT – Efficient Federated Black-box Prompt Tuning for LLMs (ICML 2024)

• ConDistFL – Conditional Distillation from Partially Annotated Data (DeCaF 2023)

• FedOBD – Opportunistic Block Dropout for Large-scale Neural Networks (IJCAI 2023)

• Fair Federated Medical Image Segmentation (CVPR 2023)

• One-shot Vertical FL with Limited Overlapping Samples (ICCV 2023)

• FedSM – Closing the Generalization Gap of Cross-silo FL (CVPR 2022)

• Quantifying Data Leakage in Gradient Inversion Attacks (IEEE TMI 2022)

• Auto-FedRL – Federated Hyperparameter Optimization (ECCV 2022)

• FedBN – Federated Learning on Non-IID Features (ICLR 2021)

• Privacy-preserving Federated Brain Tumour Segmentation (MLMI 2019)

Contributing

Interested in contributing your own federated learning research? See the research contribution guide.