Facebook Report on Privacy of fNIRS data
This addresses privacy concerns for users of fNIRS data in medical or research applications, but appears incremental as it applies existing methods like DP and federated learning to a new domain.
The project tackled the problem of privacy in fNIRS data by developing machine learning techniques with differential privacy and certified robustness in centralized and federated settings, aiming to prevent private information leakage without specifying concrete results or numbers.
The primary goal of this project is to develop privacy-preserving machine learning model training techniques for fNIRS data. This project will build a local model in a centralized setting with both differential privacy (DP) and certified robustness. It will also explore collaborative federated learning to train a shared model between multiple clients without sharing local fNIRS datasets. To prevent unintentional private information leakage of such clients' private datasets, we will also implement DP in the federated learning setting.