Neurally Integrated Finite Elements for Differentiable Elasticity on Evolving Domains
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909570775908352 |
|---|---|
| author | Daviet, Gilles Shen, Tianchang Sharp, Nicholas Levin, David I. W. |
| author_facet | Daviet, Gilles Shen, Tianchang Sharp, Nicholas Levin, David I. W. |
| contents | We present an elastic simulator for domains defined as evolving implicit functions, which is efficient, robust, and differentiable with respect to both shape and material. This simulator is motivated by applications in 3D reconstruction: it is increasingly effective to recover geometry from observed images as implicit functions, but physical applications require accurately simulating and optimizing-for the behavior of such shapes under deformation, which has remained challenging. Our key technical innovation is to train a small neural network to fit quadrature points for robust numerical integration on implicit grid cells. When coupled with a Mixed Finite Element formulation, this yields a smooth, fully differentiable simulation model connecting the evolution of the underlying implicit surface to its elastic response. We demonstrate the efficacy of our approach on forward simulation of implicits, direct simulation of 3D shapes during editing, and novel physics-based shape and topology optimizations in conjunction with differentiable rendering. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_09417 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Neurally Integrated Finite Elements for Differentiable Elasticity on Evolving Domains Daviet, Gilles Shen, Tianchang Sharp, Nicholas Levin, David I. W. Graphics Computer Vision and Pattern Recognition We present an elastic simulator for domains defined as evolving implicit functions, which is efficient, robust, and differentiable with respect to both shape and material. This simulator is motivated by applications in 3D reconstruction: it is increasingly effective to recover geometry from observed images as implicit functions, but physical applications require accurately simulating and optimizing-for the behavior of such shapes under deformation, which has remained challenging. Our key technical innovation is to train a small neural network to fit quadrature points for robust numerical integration on implicit grid cells. When coupled with a Mixed Finite Element formulation, this yields a smooth, fully differentiable simulation model connecting the evolution of the underlying implicit surface to its elastic response. We demonstrate the efficacy of our approach on forward simulation of implicits, direct simulation of 3D shapes during editing, and novel physics-based shape and topology optimizations in conjunction with differentiable rendering. |
| title | Neurally Integrated Finite Elements for Differentiable Elasticity on Evolving Domains |
| topic | Graphics Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2410.09417 |