Deep Injective Prior for Inverse Scattering

This repository is the official Tensorflow Python implementation of "Deep Injective Prior for Inverse Scattering" published in IEEE Transactions on Antennas and Propagation, 2023.

| Project Page |

Requirements

(This code is tested with tensorflow-gpu 2.3.0, Python 3.8.3, CUDA 11.0 and cuDNN 7.)

• numpy
• scipy
• matplotlib
• sklearn
• opencv-python
• tensorflow-gpu==2.3.0
• tensorflow-probability==0.11.1

Experiments

Training the injective generative model:

We used MNIST and the custom ellipses datasets. You can download the ellipses dataset from here, unzip the file, and put the .npy file in the datasets/ellipses/. You should specify the training parameters in config.py and run the following command:

python3 main.py

Solving inverse scattering problem using the injective generator as prior:

You should download the scattering configuration files from here and put "scattering_config" folder in the project directory. You should specify the parameters for MAP estimation and posterior sampling in config.py and run the following command:

python3 main.py

Citation

If you find the code or our dataset useful in your research, please consider citing the paper.

@article{khorashadizadeh2023deep,
  title={Deep Injective Prior for Inverse Scattering},
  author={Khorashadizadeh, AmirEhsan and Khorashadizadeh, Vahid and Eskandari, Sepehr and Vandenbosch, Guy AE and Dokmani{\'c}, Ivan},
  journal={IEEE Transactions on Antennas and Propagation},
  year={2023},
  publisher={IEEE}
}