Review: SMPL: A Skinned Multi-Person Linear Model¶
This is my note on the ACM 2015 paper “SMPL: A Skinned Multi-Person Linear Model” by Loper, Matthew and Mahmood, Naureen and Romero, Javier and Pons-Moll, Gerard and Black, Michael J.
Introduction¶
SMPL is a skinned vertex-based model that accurately represents a wide variety of body shapes in natural human poses.
In SMPL, the pose information is independent from the shape information.
Some parameters are learned from data:
- rest pose template
- blend weights
- pose-dependent blend shapes
- identity-dependent blend shape
- a regressor from vertices to joint locations
In SMPL, the authors learn blend shapes (identity, pose, soft-tissue dynamics) that are additively combined with a rest template before being tranformed by blend skinning.
The pose blend shapes are formulated as a linear function of the elements of the part rotation matrices.
Objective function penalizes the per-vertex disparities between registered meshses and the model.
Learn pose¶
Data: 1786 high-resolution 3D scans of different subjects in a wide variety of poses.
- align the template mesh to each scan
- optimize for:
- the blend weights,
- pose-dependent blend shapes
- mean template shape (rest pose)
- a regressor from shape to joint locations
Learn shape¶
Data: CAESAR dataset (~2000 scans per gender).
- Register a template mesh to each scan
- Pose normalize the data: which is critical when learning a vertex-based shape model.
- Use principal component analysis (PCA) to learn body shape blend shapes.