Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/maxjiang93/ugscnn

Spherical CNNs on Unstructured Grids Using Parameterized Differential Operators
https://github.com/maxjiang93/ugscnn

Last synced: 14 days ago
JSON representation

Spherical CNNs on Unstructured Grids Using Parameterized Differential Operators

Awesome Lists containing this project

README 

        
## UGSCNN: Spherical CNNs on Unstructured Grids

 

By: [Chiyu "Max" Jiang](http://maxjiang.ml/), [Jingwei Huang](http://stanford.edu/~jingweih/), [Karthik Kashinath](http://www.nersc.gov/about/nersc-staff/data-analytics-services/karthik-kashinath/), [Prabhat](http://www.nersc.gov/about/nersc-staff/data-analytics-services/prabhat/), [Philip Marcus](http://www.me.berkeley.edu/people/faculty/philip-s-marcus), [Matthias Niessner](http://niessnerlab.org/)



\[[Project Website](http://www.maxjiang.ml/proj/ugscnn)\] \[[Paper](https://openreview.net/pdf?id=Bkl-43C9FQ)\]

 

![teaser](doc/ugscnn_teaser.png "UGSCNN_teaser")



### Introduction

This repository is based on our ICLR 2019 paper: [UGSCNN: Spherical CNNs on Unstructured Grids](https://openreview.net/pdf?id=Bkl-43C9FQ). The [project webpage](http://www.maxjiang.ml/proj/ugscnn) presents an overview of the project. 



In this project, we present an alternative convolution kernel for deploying CNNs on unstructured grids, using parameterized differential operators. More specifically we evaluate this method for the spherical domain that is discretized using the icosahedral spherical mesh. Our unique convolution kernel parameterization scheme achieves high parameter efficiency compared to competing methods. We evaluate our model for classification as well as semantic segmentation tasks. Please see `experiments/` for detailed examples.



Our deep learning code base is written using [PyTorch](https://pytorch.org/) in Python 3, in conjunction with standard ML packages such as [Scikit-Learn](http://scikit-learn.org/stable/) and [Numpy](http://www.numpy.org/).



### Generate or download mesh files

To acquire the mesh files used in this project, run the provided script `gen_mesh.py`. 

```bash

python gen_mesh.py

```

To locally generate the mesh files, the [Libigl](http://libigl.github.io/libigl/) library is required. Libigl is mainly used for computing the Laplacian and Derivative matrices that are stored in the pickle files. Alternatively, the script will download precomputed pickles if the library is not available.



### Run experiments

To run experiments, please find details instructions in under individual experiments in [`experiments`](experiments). For most experiments, simply running the script `run.sh` is sufficient to start the training process:

```bash

chmod +x run.sh

./run.sh

```

The script will automatically download data files if needed.



### Citation

If you find our code useful for your work, please consider citing our paper:

```

@inproceedings{

jiang2018spherical,

title={Spherical {CNN}s on Unstructured Grids},

author={Chiyu Max Jiang and Jingwei Huang and Karthik Kashinath and Prabhat and Philip Marcus and Matthias Niessner},

booktitle={International Conference on Learning Representations},

year={2019},

url={https://openreview.net/forum?id=Bkl-43C9FQ},

}

```



### Credits

We used code from open-source repositories, including [S2CNN](https://github.com/jonas-koehler/s2cnn), [Libigl](http://libigl.github.io/libigl/), among others.



We thank Zhewen Xu from Jilin University for help in recovering the climate sphere data files (`climate_sphere_l5.zip`), and for providing a download port for scholars in mainland China via the Baidu drive [link](https://pan.baidu.com/s/11RV69AtewamTFH61gvD_8g?pwd=uq8s) (code：uq8s).



### Contact

Please contact [Max Jiang](mailto:[email protected]) if you have further questions!