Data-driven synthesis of smoke flows with CNN-based feature descriptors

By Mengyu Chu, Nils Thürey. 2017

281 citations recorded by the source corpus.

We present a novel data-driven algorithm to synthesize high resolution flow simulations with reusable repositories of space-time flow data. In our work, we employ a descriptor learning approach to encode the similarity between fluid regions with differences in resolution and numerical viscosity. We use convolutional neural networks to generate the descriptors from fluid data such as smoke density and flow velocity. At the same time, we present a deformation limiting patch advection method which allows us to robustly track deformable fluid regions. With the help of this patch advection, we generate stable space-time data sets from detailed fluids for our repositories. We can then use our learned descriptors to quickly localize a suitable data set when running a new simulation. This makes our approach very efficient, and resolution independent. We will demonstrate with several examples that our method yields volumes with very high effective resolutions, and non-dissipative small scale details that naturally integrate into the motions of the underlying flow.

Subfields: Computational Mechanics, Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Mechanics

1. Lukas Prantl
2. Nils Thürey
3. Boris Bonev
4. Suvranu De
5. Rahul
6. Zhongqing Han
7. Philipp Holl
8. Robert Brand

In Computer Graphics and Computer-Aided Design

1. Lukas Prantl
2. Nils Thürey
3. Boris Bonev
4. Suvranu De
5. Rahul
6. Zhongqing Han
7. Robert Brand
8. Philipp Holl

In Computer Vision and Pattern Recognition

1. Nils Thürey
2. Lukas Prantl
3. Boris Bonev
4. Suvranu De
5. Rahul
6. Zhongqing Han
7. Philipp Holl
8. Robert Brand

Citations in the corpus

• Unsupervised deep learning for super-resolution reconstruction of turbulence (2020, background): “For example, it is possible to account for temporally successive data by adding a discriminator that considers temporal effects (Xie et al. 2018; Kim & Lee 2020a).”
• A deep learning-based hybrid approach for the solution of multiphysics problems in electrosurgery (2019, methodology): “CNNs have been shown to be effective in the simulation of the Navier-Stokes equations [22,27].”
• A Gentle Introduction to Deep Learning in Medical Image Processing (2018, background): “So far this has been exploited in the gaming industry to compute realistically appearing physics engines [92], or for smoke simulation [93] in real-time.”
• Towards Physics-informed Deep Learning for Turbulent Flow Prediction (2019, background): “[3, 4] develop deep learning models in the context of fluid animation, 32 where being physically meaningful is less of a concern.”
• Generating Liquid Simulations with Deformation-aware Neural Networks (2017, methodology): “Our model shares the goal to learn flow physics based on examples with other methods [23, 39, 9], but in contrast to these we focus on 4D volumes of physics data, instead of localized windows at single instances of time.”
• Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers (2020, methodology): “[74] Y.”