SMPL is a learned posable model of human body shape. We show it is more accurate than BlendSCAPE trained on the same data. We also extend it to realistically model dynamic soft-tissue deformations. It is compatible with existing rendering engines and available for research purposes.

Matthew Loper, Max Planck Institute for Intelligent Systems
Naureen Mahmood, Max Planck Institute for Intelligent Systems
Javier Romero, Max Planck Institute for Intelligent Systems
Gerard Pons-Moll, Max Planck Institute for Intelligent Systems
Michael Black, Max Planck Institute for Intelligent Systems