A self-supervised learning-based approach for multi-person 3D pose estimation from multiple RGB cameras without requiring annotated datasets.
