Skip to content

← ArchivePaper2022

Motion Guided Deep Dynamic 3D Garments

Meng Zhang, Duygu Ceylan, Niloy J. Mitra

SIGGRAPH AsiaAcademic40 citesCFXML Deformation

Motion-conditioned neural model generalizes garment dynamics to unseen body shapes and motions by disentangling body and garment representations.

Abstract

Realistic dynamic garments on animated characters have many AR/VR applications. While authoring such dynamic garment geometry is still a challenging task, data-driven simulation provides an attractive alternative, especially if it can be controlled simply using the motion of the underlying character. In this work, we focus on motion guided dynamic 3D garments, especially for loose garments. In a data-driven setup, we first learn a generative space of plausible garment geometries. Then, we learn a mapping to this space to capture the motion dependent dynamic deformations, conditioned on the previous state of the garment as well as its relative position with respect to the underlying body. Technically, we model garment dynamics, driven using the input character motion, by predicting per-frame local displacements in a canonical state of the garment that is enriched with frame-dependent skinning weights to bring the garment to the global space. We resolve any remaining per-frame collisions by predicting residual local displacements. The resultant garment geometry is used as history to enable iterative roll-out prediction. We demonstrate plausible generalization to unseen body shapes and motion inputs, and show improvements over multiple state-of-the-art alternatives. Code and data is released in https://geometry.cs.ucl.ac.uk/projects/2022/MotionDeepGarment/

How to read this

Category
Method: data-driven dynamic garment model
Contributions
  • Learns a generative space of plausible garment geometries, then a motion-dependent mapping into it conditioned on prior garment state and relative body position
  • Predicts per-frame local displacements in a canonical garment state enriched with frame-dependent skinning weights, plus residual displacements to resolve collisions
  • Enables iterative roll-out prediction that generalizes to unseen body shapes and motions, targeting loose garments
Context
Builds on learning-based garment animation such as Santesteban et al. (2019) virtual try-on, extending it toward loose garments with motion-dependent dynamics.Builds on: Learning-Based Animation of Clothing for Virtual Try-On
Correctness
Demonstrated to generalize plausibly to unseen body shapes and motions, but it is data-driven and focused on loose garments; collision handling is approximate (residual displacements), so a reader should not expect hard physical guarantees.
Clarity
Moderately technical; a first pass gives the disentanglement idea, a second pass is needed for the canonical-state displacement and roll-out formulation.
How to read it
First pass for the body/garment disentanglement and roll-out scheme; second pass on the canonical displacement plus skinning-weight enrichment if you care about how dynamics are encoded.

Builds on

Built upon by

Nothing yet.

Related work

Keywords

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 →