# SnapMoGen

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

- id: `snapmogen`
- kind: language and reasoning annotation
- robot or device: human body mocap with long text (class: other)
- how collected: mocap plus manual and LLM text
- size as stated: 43.7 hours, 20,450 clips, 122,565 descriptions
- hours: 43.7 | hours only claimed: none | episodes: not stated
- year: 2025
- organisation: Chuan Guo et al. (Snap)
- licence: other (class: custom terms)
- access: open (class: open)
- link (dataset): https://huggingface.co/datasets/Ericguo5513/SnapMoGen
- Hugging Face repository: `Ericguo5513/SnapMoGen`, 569 downloads in 30 days, https://huggingface.co/datasets/Ericguo5513/SnapMoGen
- note: Descriptions average 48 words, four times HumanML3D.

Confirm the licence at the link before relying on it.