Explore high-quality datasets for your AI and machine learning projects.
A dataset that pairs surface electromyography (sEMG) recordings with hand motion capture recordings. It contains 25,253 HDF5 files, each holding time‑aligned 2 kHz sEMG data and joint angles of a single hand during a single stage. The dataset covers 193 participants, totaling 370 hours across 29 stages.
The MyoUP database contains recordings from eight complete participants (3 female, 5 male; 1 left‑handed, 7 right‑handed; age 22.38 ± 1.06 years). The acquisition process consists of three parts: five basic finger motions, twelve isometric and isotonic hand configurations, and five grasp gestures. Volunteers familiarized themselves with the protocol before each set of exercises. Participants were instructed to repeat each gesture five times, each lasting 5 seconds, with 5‑second rests to avoid muscle fatigue.