Deep Fluids: A Generative Network for Parameterized Fluid Simulations

TL;DR

This paper introduces a deep learning-based generative model for fluid simulations that can produce accurate, diverse, and fast simulations from parameterized inputs, with applications in interpolation, compression, and real-time simulation.

Contribution

It presents a novel divergence-free loss function and a deep generative network capable of modeling complex fluid behaviors efficiently.

Findings

Velocity fields generated up to 700x faster than traditional methods.

Achieved compression rates of up to 1300x.

Successfully handled complex parameterizations in reduced spaces.

Abstract

This paper presents a novel generative model to synthesize fluid simulations from a set of reduced parameters. A convolutional neural network is trained on a collection of discrete, parameterizable fluid simulation velocity fields. Due to the capability of deep learning architectures to learn representative features of the data, our generative model is able to accurately approximate the training data set, while providing plausible interpolated in-betweens. The proposed generative model is optimized for fluids by a novel loss function that guarantees divergence-free velocity fields at all times. In addition, we demonstrate that we can handle complex parameterizations in reduced spaces, and advance simulations in time by integrating in the latent space with a second network. Our method models a wide variety of fluid behaviors, thus enabling applications such as fast construction of simulations, interpolation of fluids with different parameters, time re-sampling, latent space simulations, and compression of fluid simulation data. Reconstructed velocity fields are generated up to 700x faster than re-simulating the data with the underlying CPU solver, while achieving compression rates of up to 1300x.