Multi Angle Hand Gesture EMG Dataset (MAHG-EMG)

This dataset consists of surface electromyography (sEMG) signals collected from 10 healthy individuals as they performed five unique hand gestures. Each participant carried out the gestures at five different arm positions: 0\u00b0, 45\u00b0, -45\u00b0, 90\u00b0, and -90\u00b0. The data for each gesture were captured at a sampling rate of 1259 Hz  the Delsys Trigno wireless system.Four surface electromyography (sEMG) electrodes were evenly distributed around the forearm, positioned just below the radio-humeral joint. Three electrodes were placed on the anterior (palm-facing) side, targeting the ,Flexor Carpi Radialis, Flexor Digitorum Superficialis  and Flexor Carpi Ulnaris, while a single electrode was positioned on the posterior side over the extensor digitorum. Data collection involved an initial 5-second rest period followed by 5 seconds of active gesture execution, with only movement data retained after discarding the rest phases. Additionally, a Python-based graphical user interface (GUI) was implemented to standardize the procedure, automatically displaying gesture images during data collection to provide visual guidance for participants throughout each trial. The dataset comprises 100 CSV files, corresponding to 10 repetitions per subject. Each file contains sequential data for all gestures performed in a single repetition, with labeled states indicating the specific gesture being executed. This well-structured dataset supports research in gesture classification, human-machine interaction, and biomechanical analysis offering multi-angle EMG signals in a reproducible format.