A Compact Surface EMG Dataset for Hand Gesture Recognition Utilizing a Minimal Channel Configuration on Forearm Muscles

The presented dataset includes electromyography (EMG) signals acquired by performing a set of gestures for the purpose of hand gesture recognition and human-computer interaction.

The data comprises 3-channel surface electromyography signals captured from 5 healthy participants while performing 14 different hand gestures; each gesture was performed 15 times per participant. The data collection process involved placing three sEMG sensors on three key forearm muscles; the raw 3.5s signals were acquired with a sampling rate of 1 kHz and stored in a .csv format, providing easy accessibility and visualization.

This dataset can be employed with artificial intelligence models in hand gesture classification, human-computer interaction (HCI), biomedical signal processing, and prosthetic hand control and rehabilitation, as well as validating machine learning models created with other datasets.