Hand Gesture Recognition via sEMG signals with CNNs (Electrical and Computer Engineering - MSc Thesis)

The source code for the real-time hand gesture recognition algorithm based on Temporal Muscle Activation maps of multi-channel surface electromyography (sEMG) signals (ICASSP 2021)

Accompaniment code for 'Hilbert sEMG data scanning for hand gesture recognition based on Deep Learning' published in NCAA.

A toolbox for offline muscle activity onset/offset detection in multi-channel EMG data.

Source code for multiple parameter modelling of synthetic electromyography data.

Decomposition of sEMG signals via Blind Source Separation

My work during the research internship "Automotive Control using Surface Electromyography" at The University of Tokyo, Jun.-Sep. 2020

Bayesian self-adaptive pattern recognition for sEMG

This repository contains sEMG Data of 13 subjects recorded with the Myo Armband.

Auto-learning search framework based on a weighted double Q-learning algorithm:"Integrated block-wise neural network with auto-learning search framework for finger gesture recognition using sEMG signals"

Computationally-free personalization at test time for sEMG gesture classification. Fast (gpu/cpu) ninapro API.

A comprehensive sEMG dataset recorded at Mayo Hospital Lahore and National University of Sciences & Technology. It includes raw signals from healthy subjects and stroke patients performing six upper limb gestures, captured with Myo armband following rigorous ethical standards.

Python algorithm to assess muscle activation patterns during cyclical movements

Shazam is an application used for audio segments recognition, based on digital fingerprint which are extracted from a time-frequency graph (spectrogram). In this approach, we extend its uses to sEMG signal classification .

Movement classification from sEMG signals across different subjects.

A RNN Classification of various leg movements using different lower limb muscle sEMG data.

Collection of Arduino code for my bachelorthesis which is a cybernetic prosthetic prototype of the right arm that can be controlled via surface level EMG sensors on the lower forearm.

This Dataset contains measurements of wrist flexor muscles for 4 different gestures, initial starting position was: sitting with a 90° flexion angle of elbows, hips and knees.

Blind source separation of sEMG signals on PULP platform

EMG-based Gesture Recognition for Robot Control