INSTALL.md

Installation updated (Anaconda env)

Requirements

a. Create a conda environment

conda create --name openpcdet python=3.8

b. Uninstall of old CUDA version

Follow these guides to remove previous versions of CUDA. https://gist.github.com/kmhofmann/cee7c0053da8cc09d62d74a6a4c1c5e4

https://medium.com/virtual-force-inc/a-step-by-step-guide-to-install-nvidia-drivers-and-cuda-toolkit-855c75efcdb6

With nvidia-smi you will see the max. supported version of CUDA. After you installed a proper versions verify with nvcc --version your fresh installed CUDA version.

c. Installation of OpenPCDet with Anaconda

conda activate openpcdet

Go to https://pytorch.org/ to copy the command for installing PyTorch with the corresponding CUDA version.

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

Install the SparseConv library from [spconv]. In the table you see CUDA version and the related installation command.

pip install spconv-cu118

Install

Clone the OpenPCDet repository from my account and navigate into it.

git clone https://github.com/Petros626/OpenPCDet-data-augmentor.git

cd OpenPCDet

Install this pcdet library and its dependent libraries by running the following command

pip install -r requirements.txt
python setup.py develop

Installation (old)

Requirements

All the codes are tested in the following environment:

• Linux (tested on Ubuntu 14.04/16.04/18.04/20.04/21.04)
• Python 3.6+
• PyTorch 1.1 or higher (tested on PyTorch 1.1, 1,3, 1,5~1.10)
• CUDA 9.0 or higher (PyTorch 1.3+ needs CUDA 9.2+)
• spconv v1.0 (commit 8da6f96) or spconv v1.2 or spconv v2.x

Install pcdet v0.5

NOTE: Please re-install pcdet v0.5 by running python setup.py develop even if you have already installed previous version.

a. Clone this repository.

git clone https://github.com/open-mmlab/OpenPCDet.git

b. Install the dependent libraries as follows:

• Install the SparseConv library, we use the implementation from [spconv].
  • If you use PyTorch 1.1, then make sure you install the spconv v1.0 with (commit 8da6f96) instead of the latest one.
  • If you use PyTorch 1.3+, then you need to install the spconv v1.2. As mentioned by the author of spconv, you need to use their docker if you use PyTorch 1.4+.
  • You could also install latest spconv v2.x with pip, see the official documents of spconv.

c. Install this pcdet library and its dependent libraries by running the following command:

python setup.py develop