← ArchivePaper2013

Capture and Statistical Modeling of Arm-Muscle Deformations

Thomas Neumann, Kiran Varanasi, Nils Hasler, Marcus Wacker, Markus Magnor, Christian Theobalt

Multi-camera capture and semi-parametric statistical model of shoulder-arm skin and muscle deformations across subjects and poses.

Open the paper ↗Trace in the lineage graph

Abstract

We present a comprehensive data‐driven statistical model for skin and muscle deformation of the human shoulder‐arm complex. Skin deformations arise from complex bio‐physical effects such as non‐linear elasticity of muscles, fat, and connective tissue; and vary with physiological constitution of the subjects and external forces applied during motion. Thus, they are hard to model by direct physical simulation. Our alternative approach is based on learning deformations from multiple subjects performing different exercises under varying external forces. We capture the training data through a novel multi‐camera approach that is able to reconstruct fine‐scale muscle detail in motion. The resulting reconstructions from several people are aligned into one common shape parametrization, and learned using a semi‐parametric non‐linear method. Our learned data‐driven model is fast, compact and controllable with a small set of intuitive parameters, pose, body shape and external forces, through which a novice artist can interactively produce complex muscle deformations. Our method is able to capture and synthesize fine‐scale muscle bulge effects to a greater level of realism than achieved previously. We provide quantitative and qualitative validation of our method.

How to read this

Category: Capture system + data-driven statistical model
Contributions: Multi-camera capture approach that reconstructs fine-scale muscle detail of the shoulder-arm complex in motion
Aligns reconstructions from several subjects into a common shape parametrization and learns a semi-parametric non-linear deformation model
Yields a fast, compact, controllable model driven by intuitive parameters (pose, body shape, external forces) for interactive muscle bulging
Context: An alternative to direct physical muscle simulation (Teran et al. Creating and Simulating Skeletal Muscle from the Visible Human Data Set), learning skin and muscle deformation from captured multi-subject data instead.Builds on: Creating and Simulating Skeletal Muscle from the Visible Human Data Set
Correctness: Validated as capturing fine-scale bulge effects across subjects and external forces; as a learned model its fidelity is bounded by the captured subjects, poses, and force conditions, so extrapolation beyond the training distribution is uncertain.
Clarity: Accessible; a first pass conveys the capture-then-learn pipeline and the control parameters, a second pass clarifies the alignment and the semi-parametric fitting.
How to read it: Read for the capture rig and the parametrization/learning choices; a second pass on the statistical model pays off if you want to drive or extend muscle deformation from data.

Builds on

Creating and Simulating Skeletal Muscle from the Visible Human Data Set 2005

Built upon by

Nothing yet.

Related work

Data-driven Modeling of Skin and Muscle Deformation 2008 / SIGGRAPH
Anatomically Based Modeling 1997 / SIGGRAPH
How to Build a Human: Practical Physics-Based Character Animation 2016 / DigiPro
A Neural Network Model for Efficient Musculoskeletal-Driven Skin Deformation 2024 / SIGGRAPH

Keywords

muscle deformation data-driven model arm-muscle statistical modeling skin deformation multi-view reconstruction mocap kernel ridge regression body shape external forces pose space deformation

This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →