Code for the papers,
- CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing.
- Gabor is Enough: Interpretable Deep Denoising with a Gabor Synthesis Dictionary Prior, [ieee]
Install packages listed in reqiurements.txt
to replicate the environment.
Edit the example args.json
file to define your own model and training.
Architecture hyperparameter notation follows the journal article (K -
iterations, M - subbands, C - input-channels, P - filter side-length, etc.).
Paramters in args['model']
will be passed directly to the constructor of the
network, defined in model/net.py
. The option args['model_type']
can be
['CDLNet', 'GDLNet', 'JDD_CDLNet', 'DnCNN', 'FFDNet']
. Note that DnCNN and FFDNet implementations are provided but were not used to generate the numbers reported in the paper (see KAIR).
Once your args file is defined, training may be performed via
$ python train.py path/to/args.json
Testing of the trained model is done via
$ python analyze.py path/to/args.json --test path/to/dataset/ --noise_level 25
where the checkpoint referenced in args['paths']['ckpt']
will be loaded.
Additional analysis of trained models can be performed with other command-line
options (see $ python analyze.py -h
).
website, supplementary material
If you find this code/work useful, please cite us:
@article{janjusevicCDLNet2022,
author={Janjušević, Nikola and Khalilian-Gourtani, Amirhossein and Wang, Yao},
journal={IEEE Open Journal of Signal Processing},
title={{CDLNet}: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing},
year={2022},
volume={3},
number={},
pages={196-211},
doi={10.1109/OJSP.2022.3172842}
}
@INPROCEEDINGS{janjusevicGDLNet2022,
author={Janjušević, Nikola and Khalilian-Gourtani, Amirhossein and Wang, Yao},
booktitle={2022 IEEE 14th Image, Video, and Multidimensional Signal Processing Workshop (IVMSP)},
title={Gabor is Enough: Interpretable Deep Denoising with a Gabor Synthesis Dictionary Prior},
year={2022},
volume={},
number={},
pages={1-5},
doi={10.1109/IVMSP54334.2022.9816313}
}