emg2pose

From the catalogue. Listed by one researcher from one source page.

• id: emg2pose
• kind: humanoid and human motion
• robot or device: 16 channel sEMG wristband with 26-camera motion capture hand pose labels (class: sensor only)
• how collected: 193 users performed gesture stages wearing sEMG bands in a mocap rig
• size as stated: 193 users, 29 stages, 2 kHz 16 channel sEMG, 431 GiB; 370 hours
• hours: 370 | hours only claimed: none | episodes: not stated
• year: 2024
• organisation: Sasha Salter et al. (facebookresearch)
• licence: CC-BY-NC-SA-4.0 (class: non commercial)
• access: open download (class: open)
• link (dataset): https://github.com/facebookresearch/emg2pose
• note: Wearable sEMG and hand pose, no egocentric video.

Confirm the licence at the link before relying on it.