This repo contains the official implementation for the paper SNIPS: Solving Noisy Inverse Problems Stochastically.

by Bahjat Kawar, Gregory Vaksman, and Michael Elad, Computer Science Department, Technion.

Run the following conda line to install all necessary python packages for our code and set up the snips environment.

conda env create -f environment.yml

The environment includes cudatoolkit=11.0. You may change that depending on your hardware.

main.py is the file that you should run for both training and sampling. Execute python main.py --help to get its usage description:

usage: main.py [-h] --config CONFIG [--seed SEED] [--exp EXP] --doc DOC
               [--comment COMMENT] [--verbose VERBOSE] [-i IMAGE_FOLDER]
               [-n NUM_VARIATIONS] [-s SIGMA_0] [--degradation DEGRADATION]

optional arguments:
  -h, --help            show this help message and exit
  --config CONFIG       Path to the config file
  --seed SEED           Random seed
  --exp EXP             Path for saving running related data.
  --doc DOC             A string for documentation purpose. Will be the name
                        of the log folder.
  --comment COMMENT     A string for experiment comment
  --verbose VERBOSE     Verbose level: info | debug | warning | critical
  -i IMAGE_FOLDER, --image_folder IMAGE_FOLDER
                        The folder name of samples
  -n NUM_VARIATIONS, --num_variations NUM_VARIATIONS
                        Number of variations to produce
  -s SIGMA_0, --sigma_0 SIGMA_0
                        Noise std to add to observation
  --degradation DEGRADATION
                        Degradation: inp | deblur_uni | deblur_gauss | sr2 |
                        sr4 | cs4 | cs8 | cs16

Configuration files are in config/. You don't need to include the prefix config/ when specifying --config . All files generated when running the code is under the directory specified by --exp. They are structured as:

<exp> # a folder named by the argument `--exp` given to main.py
├── datasets # all dataset files
│   ├── celeba # all CelebA files
│   └── lsun # all LSUN files
├── logs # contains checkpoints and samples produced during training
│   └── <doc> # a folder named by the argument `--doc` specified to main.py
│      └── checkpoint_x.pth # the checkpoint file saved at the x-th training iteration
├── image_samples # contains generated samples
│   └── <i>
│       ├── stochastic_variation.png # samples generated from checkpoint_x.pth, including original, degraded, mean, and std   
│       ├── results.pt # the pytorch tensor corresponding to stochastic_variation.png
│       └── y_0.pt # the pytorch tensor containing the input y of SNIPS

You can download the aligned and cropped CelebA files from their official source here. The LSUN files can be downloaded using this script. For our purposes, only the validation sets of LSUN bedroom and tower need to be downloaded.

If we want to run SNIPS on CelebA for the problem of super resolution by 2, with added noise of standard deviation 0.1, and obtain 3 variations, we can run the following

python main.py -i celeba --config celeba.yml --doc celeba -n 3 --degradation sr2 --sigma_0 0.1

Samples will be saved in <exp>/image_samples/celeba.

The available degradations are: Inpainting (inp), Uniform deblurring (deblur_uni), Gaussian deblurring (deblur_gauss), Super resolution by 2 (sr2) or by 4 (sr4), Compressive sensing by 4 (cs4), 8 (cs8), or 16 (cs16). The sigma_0 can be any value from 0 to 1.

Link: https://drive.google.com/drive/folders/1217uhIvLg9ZrYNKOR3XTRFSurt4miQrd?usp=sharing

These checkpoint files are provided as-is from the authors of NCSNv2. You can use the CelebA, LSUN-bedroom, and LSUN-tower datasets' pretrained checkpoints. We assume the --exp argument is set to exp.

This repo is largely based on the NCSNv2 repo, and uses modified code from this repo for implementing the blurring matrix.

If you find the code/idea useful for your research, please consider citing

@article{kawar2021snips,
  title={{SNIPS}: Solving noisy inverse problems stochastically},
  author={Kawar, Bahjat and Vaksman, Gregory and Elad, Michael},
  journal={Advances in Neural Information Processing Systems},
  volume={34},
  pages={21757--21769},
  year={2021}
}