InteFL Research Framework

InteFL is a framework for designing and running federated learning (FL) experiments, focused on anomaly detection, adversarial robustness and threat mitigation in distributed learning environments. It is used to detect malicious client behavior and model poisoning attacks, and it supports student capstone projects on FL security at RIT’s Data Quality Lab.

I have built a foundation for the framework with my colleagues at Data Quality Lab. Currently, I continue to develop and maintain the framework, which includes development planning, code reviewing and feedback for participating collaborators. I have technically advised several graduate capstone projects focused on extending framework functionality.