Complex Signal Denoising and Interference Mitigation for Automotive Radar Using Convolutional Neural Networks
This repository contains the code for training and evaluating CNNs for automotive radar signal denoising as introduced in the paper Complex Signal Denoising and Interference Mitigation for Automotive Radar Using Convolutional Neural Networks
If you find this approach useful in your research, please consider citing:
@INPROCEEDINGS{Rock1907:Complex,
AUTHOR="Johanna Rock and Mate Toth and Elmar Messner and Paul Meissner and Franz Pernkopf",
TITLE="Complex Signal Denoising and Interference Mitigation for Automotive Radar Using Convolutional Neural Networks",
BOOKTITLE="2019 22nd International Conference on Information Fusion (FUSION) (FUSION 2019)",
YEAR=2019
}
- Clone this repository:
git clone https://github.com/johanna-rock/im_ricnn.git - Create a virtual environment from the included environment.yml and activate it
- Create using conda:
conda env create -f environment.yml - Activate using conda:
conda activate im-cnn-env
- Create using conda:
- Set the python path using
export PYTHONPATH="/path/to/imRICnn" - Download a reduced data set of simulated radar examples (200 training, 25 validation, 25 test) from https://cloud.tugraz.at/index.php/s/6GEjJaGcf7eKp2B, unzip the file and save it to
imRICnn/data/radar-data. The full data set of simulated examples is now also available, you can download it from https://nextcloud.spsc.tugraz.at/s/2qEEP6D2rHtFX64.
Run python -m run_scripts.run_training.py to train and evaluate a CNN with the configuration specified in run_training.py.
Run python -m run_scripts.run_evaluation.py to evaluate a pre-trained model with the configuration specified in run_evaluation.py.