ICLR best paper: Spherical CNNs
论文链接:https://arxiv.org/abs/1801.10130
GITHUB地址:https://github.com/zhixuanli/s2cnn
中文讲解地址:
简介:
In this paper we introduce the building blocks for constructing spherical CNNs. We propose a definition for the spherical cross-correlation that is both expressive and rotation-equivariant. The spherical correlation satisfies a generalized Fourier theorem, which allows us to compute it efficiently using a generalized (non-commutative) Fast Fourier Transform (FFT) algorithm. We demonstrate the computational efficiency, numerical accuracy, and effectiveness of spherical CNNs applied to 3D model recognition and atomization energy regression.
下面是详细的代码配置过程:
1. 运行环境配置
1.1 安装Anaconda
下载地址:https://www.anaconda.com/download/
安装参考:致Python初学者:Anaconda入门使用指南
注意要将anaconda添加到环境变量中,即在/home/yourname目录下,输入(如果你在使用bash就输入下面的命令,否则改成相应的如.zshrc):
vim .bashrc
然后在文件末尾添加上:
export PATH=/home/yourname/anaconda3/bin:$PATH
再输入以下命令使环境变量立刻生效:
source .bashrc
(非必需)
安装虚拟环境,并切换至虚拟环境,参考:https://segmentfault.com/a/1190000005828284
1.2 安装Pytorch
请选择适合自己的安装方法,这里我们选择了运行以下命令:
conda install pytorch torchvision cuda91 -c pytorch
1.3 安装CUPY
地址:https://github.com/cupy/cupy
安装方法:
pip install cupy --user或者 conda install cupy
1.4 安装lie_learn
地址:https://github.com/AMLab-Amsterdam/lie_learn.git
输入以下命令:
git clone https://github.com/AMLab-Amsterdam/lie_learn.git
python setup.py install
中间需要从Google Drive获取J_dense_0-278.npy,嗯……自己想办法吧
1.5 安装pynvrtc
输入命令:
pip install pynvrtc --user
2. 安装Spherical CNNs
切换至s2cnn文件夹下,执行:
python setup.py install
原文地址:https://www.cnblogs.com/QingHuan/p/8903459.html