High-density surface electromyography and quaternion-based hand orientation

This dataset contains synchronised recordings of high-density surface electromyography (HD-sEMG) and hand kinematics collected during repeated hand gesture execution. Dataset overview Participants: 28 subjects Recorded modalities: HD-sEMG and hand motion data Target movements: repeated hand gestures EMG acquisition: Bioelettronica Quattrocento system Hand tracking: IMU-based sensor glove Hand representation: joint and bone orientations represented using quaternions Additional motion data: linear acceleration measurements from the glove Synchronisation: EMG and glove recordings were aligned using their timestamps Data organisation The recordings are stored in HDF5 files containing participant metadata and synchronised physiological and kinematic signals. Each recording includes: filtered HD-sEMG signals (1024 Hz); gesture labels (21 gestures); repetition labels (5 repeats); EMG timestamps; bone orientation quaternions; finger orientation quaternions; linear acceleration data; glove timestamps. Participant metadata The files additionally contain: participant identifier; age; gender; dominant hand; recording date. Potential applications The dataset can be used for research in: hand gesture recognition; EMG-based human–machine interfaces; prosthetic and robotic hand control; continuous hand-pose estimation from muscle activity; multimodal modelling of neuromuscular and kinematic signals; machine-learning and deep-learning methods for decoding hand movements. The dataset is intended to support the development and evaluation of methods linking forearm muscle activity with hand movement and spatial orientation. A detailed description will be provided in a future article.