FedAssist: Federated Learning in AI-Powered Prosthetics for Sustainable and Collaborative Learning.
Journal:
Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference
PMID:
40039020
Abstract
This paper explores the integration of federated learning in developing deep learning-powered surface electromyography decoding methods for AI-controlled prosthetics. Our proposed FL framework, FedAssist, aims to preserve data ownership while fostering decentralized collaborative modeling. Specifically, it focuses on mitigating the non-independent and identically distributed (non-IID) nature of sEMG datasets. Through collaborative local-level and global-level warm-start strategies, FedAssist achieves superior performance in non-IID scenarios compared to conventional learning paradigms. This research contributes to advancing decentralized machine learning approaches in the context of sEMG, with potential applications to improve prosthetic precision and rehabilitation effectiveness.