Pointcloud_Denoising_Snowy

Implementation of the topic "RNN-based Point-Cloud Denoising in Snowy Weather Conditions"

Installation

Create a conda environment

conda create -n pc_denoising python=3.8 pip=23.0.1
conda activate pc_denoising

Install dependencies for MinkowskiEngine

conda install openblas-devel -c anaconda
conda install pytorch=1.9.0 torchvision cudatoolkit=11.1 -c pytorch -c nvidia

Install MinkowskiEngine

# Replace 11.4 with CUDA version on your machine
export CUDA_HOME=/usr/local/cuda-11.4

pip install -U git+https://github.com/NVIDIA/MinkowskiEngine -v --no-deps --install-option="--blas_include_dirs=${CONDA_PREFIX}/include" --install-option="--blas=openblas"

Install pc_denoising package and dependencies

git clone https://github.com/ANDA-Researchers/Pointcloud_Denoising_Snowy.git
cd Pointcloud_Denoising_Snowy
pip install -e .

Usage

Example of training the U-Net noise segmentation model

unet fit --config=logs/0/fit/version_0/config.yaml