SMPL

TLDR

The authors introduce SMPL, a learned, skinned vertex‑based model that captures pose‑dependent shape variation more accurately than prior models and can be used in existing graphics pipelines. SMPL is trained from thousands of aligned 3D meshes using a linear blend‑skin formulation with pose‑dependent blend shapes, identity‑dependent blend shapes, and a vertex‑to‑joint regressor, enabling accurate representation of diverse body shapes in natural poses. Evaluation shows SMPL variants using linear or dual‑quaternion blend skinning outperform a Blend‑SCAPE model trained on the same data, are compatible with rendering engines, and are released for research use.

Abstract

We present a learned model of human body shape and pose-dependent shape variation that is more accurate than previous models and is compatible with existing graphics pipelines. Our Skinned Multi-Person Linear model (SMPL) is a skinned vertex-based model that accurately represents a wide variety of body shapes in natural human poses. The parameters of the model are learned from data including the rest pose template, blend weights, pose-dependent blend shapes, identity-dependent blend shapes, and a regressor from vertices to joint locations. Unlike previous models, the pose-dependent blend shapes are a linear function of the elements of the pose rotation matrices. This simple formulation enables training the entire model from a relatively large number of aligned 3D meshes of different people in different poses. We quantitatively evaluate variants of SMPL using linear or dual-quaternion blend skinning and show that both are more accurate than a Blend-SCAPE model trained on the same data. We also extend SMPL to realistically model dynamic soft-tissue deformations. Because it is based on blend skinning, SMPL is compatible with existing rendering engines and we make it available for research purposes.