[TOG/SIGGRAPH Asia 2023] Fluid Simulation on Neural Flow Maps

by Yitong Deng, Hong-Xing Yu, Diyang Zhang, Jiajun Wu, and Bo Zhu.

Our paper and video results can be found at our project website.

Installation

Our code is tested on Windows 11 with CUDA 11.8, Python 3.10.9, PyTorch 2.0.1, and Taichi 1.6.0.

To set up the environment, first create a conda environment:

conda create -n "nfm_env" python=3.10.9 ipython
conda activate nfm_env

Then, install PyTorch with:

python -m pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu118

Finally, install the requirements with:

pip install -r requirements.txt

Simulation

For running simulation, simply execute:

python run.py

Hyperparameters can be tuned by changing the values in the file hyperparameters.py. Checkpointing is available by setting the from_frame variable to the desired frame, given that the checkpoint of that frame can be found in logs/[exp_name]/ckpts.

Visualization

The results will be stored in logs/[exp_name]/vtks. We recommend using ParaView to load these .vti files as a sequence and visualize them by selecting Volume in the Representation drop-down menu.

Bibliography

If you find our paper or code helpful, consider citing:

@article{deng2023neural,
title={Fluid Simulation on Neural Flow Maps},
author={Yitong Deng and Hong-Xing Yu and Diyang Zhang and Jiajun Wu and Bo Zhu},
journal={ACM Trans. Graph.},
volume={42},
number={6},
article={},
year={2023},
}