This code provides a Python implementation of Stein ICP presented in the paper:

Fahira Afzal Maken, Fabio Ramos, Lionel Ott, "Stein ICP for Uncertainty Estimation in Point Cloud Matching", RAL, 2021.

The code requires an NVidia GPU supported by PyTorch 1.6.0 or newer. A requirements.txt file is provided which contains the PIP installable requirements.

To facilitate the overall install a shell script pip_install.sh is provided.

All required packages can be installed by executing the pip_install.sh script.

The code is provided in the form of a Jupyter notebook, as such running the code is achieved by starting the notebook via:

jupyter notebook

The code comes with two example point clouds which by default will be aligned using Stein ICP. To align different point clouds simply modify the source and target cloud paths.