Getting Started

Installation

The DUNEdn package is open source and available at https://github.com/marcorossi5/DUNEdn.git

The package can be installed manually:

git clone https://github.com/marcorossi5/DUNEdn.git DUNEdn
cd DUNEdn
pip install .

or if you are planning to extend or develop code replace last command with:

pip install -e .

Motivation

ProtoDUNE Single Phase (SP, Technical Design Report) detector collects data containing inherent noise. This package implements deep learning models to denoise ProtoDUNE-SP events.

Neural Networks

The package, as described in the paper, implements two kind of networks for the denoising task:

All the models are implemented with the help of the PyTorch library.

The networks support also exporting to ONNX format for device agnostic inference via ONNX Runtime. More information at the relevant page:

DUNEdn models to Onnx format

How to cite DUNEdn?

When using this software in your research, please cite the following publication:

Bibtex:

@software{Rossi_DUNEdn_2022,
  author = {Rossi, Marco},
  license = {GPL-3.0},
  month = {6},
  title = {{DUNEdn}},
  url = {https://github.com/marcorossi5/DUNEdn},
  version = {2.0.0},
  year = {2022}
}

How to contribute?

For more information on how to contribute, email the author at marco.rossi@cern.ch.

FAQ

Refer to examples in the GitHub folder for more information.