Sample Dataset and Trained Model Parameters for Back-Projection Diffusion

We have uploaded a sample dataset for training and testing Back-Projection Diffusion. Trained model parameters for the dataset are also provided in tmp.zip.

For a formal description of the dataset, please refer to our preprint:

Borong Zhang, Martín Guerra, Qin Li, and Leonardo Zepeda-Núñez. "Back-Projection Diffusion: Solving the Wideband Inverse Scattering Problem with Diffusion Models." arXiv preprint arXiv:2408.02866 (2024).

In 10hsquares_trainingdata and 10hsquares_testdata, perturbations are stored as eta.h5 with the following structure:

eta.h5/      ├── /eta

The scattering data are stored as scatter.h5, or as scatter_order_n.h5 (n indicates the order of the stencil used for data generation) with the following structure:

scatter.h5/      ├── /scatter_imag_freq_1      ├── /scatter_real_freq_1      ├── /scatter_imag_freq_2      ├── /scatter_real_freq_2      ├── /scatter_imag_freq_3      ├── /scatter_real_freq_3

The tmp folder contains the trained model parameters.

For usage instructions, please refer to our GitHub repository:

https://github.com/borongzhang/back_projection_diffusion

If this dataset is useful to your research, please cite our preprint:@misc{zhang2024backprojectiondiffusionsolvingwideband,      title={Back-Projection Diffusion: Solving the Wideband Inverse Scattering Problem with Diffusion Models},       author={Borong Zhang and Martín Guerra and Qin Li and Leonardo Zepeda-Núñez},      year={2024},      eprint={2408.02866},      archivePrefix={arXiv},      primaryClass={cs.LG},      url={https://arxiv.org/abs/2408.02866}, }