Home

Awesome

Deep Learning for image deconvolution in the presence of kernel/model uncertainty

This repo provides pre-trained models and the results on benchmark datasets of our CVPR 2020 paper. main paper, supp, poster

Usage

Download pretrained models and data. Put them into separate folders.

Run test.py for deblurred images.

You can also test your data, but you need to keep the same noise level as the same.

Results

Results on True Images

Results on benchmark datasets

You can also download the deblurred results and run compute_metrics.py to compute the PSNR/SSIM with the same settings as ours.

Our results contain both synthetic blurry images with kernel error and real images.

For synthetic images, we provide the results with two settings:

For real images, please see Lai_Real from deblurred results.

Key References

Krishnan, Dilip, and Rob Fergus. "Fast image deconvolution using hyper-laplacian priors." Advances in neural information processing systems. 2009.

Zoran, Daniel, and Yair Weiss. "From learning models of natural image patches to whole image restoration." 2011 International Conference on Computer Vision. IEEE, 2011.

Kruse, Jakob, Carsten Rother, and Uwe Schmidt. " Learning to push the limits of efficient FFT-based image deconvolution. " Proceedings of the IEEE International Conference on Computer Vision. 2017.

Zhang, Kai, et al. "Learning deep CNN denoiser prior for image restoration." Proceedings of the IEEE conference on computer vision and pattern recognition. 2017.

Zhang, Jiawei, et al. "Learning fully convolutional networks for iterative non-blind deconvolution." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2017.

Ji, Hui, and Kang Wang. "Robust image deblurring with an inaccurate blur kernel." IEEE Transactions on Image processing 21.4 (2011): 1624-1634.

Whyte, Oliver, Josef Sivic, and Andrew Zisserman. "Deblurring shaken and partially saturated images." International journal of computer vision 110.2 (2014): 185-201.

Vasu, Subeesh, Venkatesh Reddy Maligireddy, and A. N. Rajagopalan. "Non-blind deblurring: Handling kernel uncertainty with CNNs." Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018.

Bibtex

@InProceedings{Nan_2020_CVPR,
author = {Nan, Yuesong and Ji, Hui},
title = {Deep Learning for Handling Kernel/model Uncertainty in Image Deconvolution},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}