NeuroFluid: Fluid Dynamics Grounding with Particle-Driven Neural Radiance Fields

Shanyan Guan, Huayu Deng, Yunbo Wang, Xiaokang Yang
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:7919-7929, 2022.

Abstract

Deep learning has shown great potential for modeling the physical dynamics of complex particle systems such as fluids. Existing approaches, however, require the supervision of consecutive particle properties, including positions and velocities. In this paper, we consider a partially observable scenario known as fluid dynamics grounding, that is, inferring the state transitions and interactions within the fluid particle systems from sequential visual observations of the fluid surface. We propose a differentiable two-stage network named NeuroFluid. Our approach consists of (i) a particle-driven neural renderer, which involves fluid physical properties into the volume rendering function, and (ii) a particle transition model optimized to reduce the differences between the rendered and the observed images. NeuroFluid provides the first solution to unsupervised learning of particle-based fluid dynamics by training these two models jointly. It is shown to reasonably estimate the underlying physics of fluids with different initial shapes, viscosity, and densities.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-guan22a, title = {{N}euro{F}luid: Fluid Dynamics Grounding with Particle-Driven Neural Radiance Fields}, author = {Guan, Shanyan and Deng, Huayu and Wang, Yunbo and Yang, Xiaokang}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {7919--7929}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/guan22a/guan22a.pdf}, url = {https://proceedings.mlr.press/v162/guan22a.html}, abstract = {Deep learning has shown great potential for modeling the physical dynamics of complex particle systems such as fluids. Existing approaches, however, require the supervision of consecutive particle properties, including positions and velocities. In this paper, we consider a partially observable scenario known as fluid dynamics grounding, that is, inferring the state transitions and interactions within the fluid particle systems from sequential visual observations of the fluid surface. We propose a differentiable two-stage network named NeuroFluid. Our approach consists of (i) a particle-driven neural renderer, which involves fluid physical properties into the volume rendering function, and (ii) a particle transition model optimized to reduce the differences between the rendered and the observed images. NeuroFluid provides the first solution to unsupervised learning of particle-based fluid dynamics by training these two models jointly. It is shown to reasonably estimate the underlying physics of fluids with different initial shapes, viscosity, and densities.} }
Endnote
%0 Conference Paper %T NeuroFluid: Fluid Dynamics Grounding with Particle-Driven Neural Radiance Fields %A Shanyan Guan %A Huayu Deng %A Yunbo Wang %A Xiaokang Yang %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-guan22a %I PMLR %P 7919--7929 %U https://proceedings.mlr.press/v162/guan22a.html %V 162 %X Deep learning has shown great potential for modeling the physical dynamics of complex particle systems such as fluids. Existing approaches, however, require the supervision of consecutive particle properties, including positions and velocities. In this paper, we consider a partially observable scenario known as fluid dynamics grounding, that is, inferring the state transitions and interactions within the fluid particle systems from sequential visual observations of the fluid surface. We propose a differentiable two-stage network named NeuroFluid. Our approach consists of (i) a particle-driven neural renderer, which involves fluid physical properties into the volume rendering function, and (ii) a particle transition model optimized to reduce the differences between the rendered and the observed images. NeuroFluid provides the first solution to unsupervised learning of particle-based fluid dynamics by training these two models jointly. It is shown to reasonably estimate the underlying physics of fluids with different initial shapes, viscosity, and densities.
APA
Guan, S., Deng, H., Wang, Y. & Yang, X.. (2022). NeuroFluid: Fluid Dynamics Grounding with Particle-Driven Neural Radiance Fields. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:7919-7929 Available from https://proceedings.mlr.press/v162/guan22a.html.

Related Material