Eva-3M

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

• id: eva-3m
• kind: egocentric
• robot or device: head-mounted device (class: human egocentric)
• how collected: body pose capture with keypoint visibility labels
• size as stated: 3.0M+ frames; 435K visibility-annotated; 27.8 hours, worked out from frames and frame rate
• hours: 27.8 | hours only claimed: none | episodes: not stated
• year: 2026
• organisation: Peng Dai et al.
• licence: not stated (class: not stated)
• access: not stated (class: not stated)
• link (paper): https://arxiv.org/abs/2602.23618
• note: Visibility-aware pose.

Confirm the licence at the link before relying on it.