English

Learning Vortex Dynamics for Fluid Inference and Prediction

Machine Learning 2023-03-17 v3 Computer Vision and Pattern Recognition Graphics

Abstract

We propose a novel differentiable vortex particle (DVP) method to infer and predict fluid dynamics from a single video. Lying at its core is a particle-based latent space to encapsulate the hidden, Lagrangian vortical evolution underpinning the observable, Eulerian flow phenomena. Our differentiable vortex particles are coupled with a learnable, vortex-to-velocity dynamics mapping to effectively capture the complex flow features in a physically-constrained, low-dimensional space. This representation facilitates the learning of a fluid simulator tailored to the input video that can deliver robust, long-term future predictions. The value of our method is twofold: first, our learned simulator enables the inference of hidden physics quantities (e.g., velocity field) purely from visual observation; secondly, it also supports future prediction, constructing the input video's sequel along with its future dynamics evolution. We compare our method with a range of existing methods on both synthetic and real-world videos, demonstrating improved reconstruction quality, visual plausibility, and physical integrity.

Keywords

Cite

@article{arxiv.2301.11494,
  title  = {Learning Vortex Dynamics for Fluid Inference and Prediction},
  author = {Yitong Deng and Hong-Xing Yu and Jiajun Wu and Bo Zhu},
  journal= {arXiv preprint arXiv:2301.11494},
  year   = {2023}
}

Comments

ICLR 2023, project webpage: https://yitongdeng.github.io/vortex_learning_webpage/

R2 v1 2026-06-28T08:22:39.079Z