SnapMoGen: A Large-scale 3D Human Motion Dataset with Expressive Text Annotations
1. Data Description
SnapMoGen is a large-scale dataset of 3D human motion sequences paired with highly expressive textual annotations.
The dataset contains 43.7 hours of high-quality motion capture (mocap) data recorded at 30 FPS. It includes a total of 20,450 motion clips,
each ranging from 4 to 12 seconds in duration. Every motion clip is accompanied by 6 textual descriptions—2 manually… See the full description on the dataset page: https://huggingface.co/datasets/Ericguo5513/SnapMoGen.
