Repository of code used to generate simulated HD-sEMG data, train CNN models using that data and evaluate the models to non-stationary signals.
This repository only contains the raw code, by inlcuding a MUAP template file and fiber location file you can run the rest of the code to generate your own data, models and evaluate those models.
This code was made to accompany the Biomedical Engineering MEng Final Year Individual Project: "Comparing Novel Convolutional Neural Networks for Real-Time Decomposition of Non-Stationary High-Density Surface EMG Signals for VR-Based Post-Stroke Rehabilitation"
By Fotios Spyridakos
Supervisor: Professor Dario Farina