N-UCLA

N-UCLA dataset is a widely used skeleton-based action recognition dataset, containing 1494 video clips featuring 10 volunteers.

BibTex: