Personalized Federated Learning For Gesture Classification

Shared repo for Kai, Ben, and Chloe, Spring 2024.

Goal is to cluster users/gestures via PCA (and potentially more advanced methods) according to gesture, user, and/or ability level. Longer term project is to develop a cluster-based personalized federated learning algorithm, and validate performance in real-time user study.

Datasets

• EMG+IMU data collected from 26 participants completing personalized gesture development tasks.