Federated Learning in Offline and Online EMG Decoding: A Privacy and Performance Perspective

TL;DR

This paper explores applying federated learning to neural interface decoding with electromyography, highlighting its privacy benefits and challenges in real-time, online settings through systematic offline and online evaluations.

Contribution

It introduces a conceptual framework for federated learning in neural interfaces and systematically evaluates its performance and privacy trade-offs in offline and online EMG decoding.

Findings

FL can improve privacy and performance offline

Standard FL assumptions do not hold in real-time online neural decoding

Performance and privacy trade-offs are more complex in online settings

Abstract

Neural interfaces offer a pathway to intuitive, high-bandwidth interaction, but the sensitive nature of neural data creates significant privacy hurdles for large-scale model training. Federated learning (FL) has emerged as a promising privacy-preserving solution, yet its efficacy in real-time, online neural interfaces remains unexplored. In this study, we 1) propose a conceptual framework for applying FL to the distinct constraints of neural interface application and 2) provide a systematic evaluation of FL-based neural decoding using high-dimensional electromyography (EMG) across both offline simulations and a real-time, online user study. While offline results suggest that FL can simultaneously enhance performance and privacy, our online experiments reveal a more complex landscape. We found that standard FL assumptions struggle to translate to real-time, sequential interactions with human-decoder co-adaptation. Our results show that while FL retains privacy advantages, it introduces performance tensions not predicted by offline simulations. These findings identify a critical gap in current FL methodologies and highlight the need for specialized algorithms designed to navigate the unique co-adaptive dynamics of sequential-user neural decoding.