This is the code repo for the paper "Provable Probabilistic Imaging using Score-based Generative Priors" (IEEE TCI | arXiv | webpage).
Setup the environment.
conda env create --file pmc.yml
Activate the environment.
conda activate pmc
Deactivate the environment.
conda deactivate
The pre-trained checkpoints are available here. Note that the checkpoints need to be placed in the proper fold for loading.
Create the folder
mkdir score_ckpt
Move the file to the folder
mv dir/to/ckpt ./score_ckpt
To run the code, simply type
python run_pmc.py -c dir/to/config.yml
Please add or configurate the data loader (pmc/test_datasets/
) to allow proper loading of your own data.
@ARTICLE{10645293,
author={Sun, Yu and Wu, Zihui and Chen, Yifan and Feng, Berthy T. and Bouman, Katherine L.},
journal={IEEE Transactions on Computational Imaging},
title={Provable Probabilistic Imaging Using Score-Based Generative Priors},
year={2024},
volume={10},
number={},
pages={1290-1305},
doi={10.1109/TCI.2024.3449114}}