Skip to content

DeepSDF Preprocessing setup

Jump to bottom
Freephi edited this page Jan 14, 2021 · 3 revisions

Locally

Setup 1)

OS: Ubuntu 18.04

GPU: GeForce GT 630

NVIDIA driver: 390.141

CUDA version: N/A

Python version: 3.6.9

pytorch version: 1.7.1+cpu

Setup 2)

OS:

NVIDIA driver:

CUDA version:

Python version:

pytorch version:

Google Cloud VM

Setup

Machine type: n1-standard-8 (8 vCPUs, 30 GB memory)

OS: Ubuntu 18.04

GPU: 1 x NVIDIA Tesla T4

NVIDIA driver:

CUDA version:

Python version: 3.6.9

pytorch version:

Installation procedure

  1. Clone DeepSDF

  2. Install requirements:

Apt install

  • git
  • python3-pip
  • libgl1-mesa-dev
  • libglew-dev
  • libpython3.6-dev
  • pkg-config
  • libeigen3-dev
  • cmake
  • zlib1g-dev
  • libgtest-dev
  • build-essential

Pip install

  • opencv-python
  • shutil
  • numpy
  • Cython
  • scipy
  • pandas
  • tensorboardX
  • munkres
  • torch -> depends on setup
  • pyopengl
  • Pillow
  • pybind11
  1. Clone DeepSDF

git clone git@github.com:facebookresearch/DeepSDF.git

cd DeepSDF

git submodule update --init

  1. Install CLI11

git clone https://github.com/CLIUtils/CLI11

cd CLI11/

mkdir build

cd build/

git submodule update --init

cmake ..

cmake --build .

sudo make install

  1. Install nanoflann

git clone https://github.com/jlblancoc/nanoflann.git

cd nanoflann/

mkdir build

cd build/

git submodule update --init

cmake ..

make -j

sudo make install

sudo mkdir /usr/local/include/nanoflann

sudo mv /usr/local/include/nanoflann.hpp /usr/local/include/nanoflann/

  1. Install Pangolin

git clone https://github.com/stevenlovegrove/Pangolin.git

cd Pangolin/

mkdir build

cd build

cmake ..

cmake --build .

sudo make install

sudo cp -r Pangolin/include/mpark/ /usr/local/include/

  1. Build DeepSDF

cd DeepSDF

mkdir build

cmake ..

make -j

sudo make install

Clone this wiki locally