/DoubleDIP

Official implementation of the paper "Double-DIP: Unsupervised Image Decomposition via Coupled Deep-Image-Priors"

Primary LanguagePython

Double Dip

Official implementation of the paper "Double-DIP": Unsupervised Image Decomposition via Coupled Deep-Image-Priors.

Paper: http://www.wisdom.weizmann.ac.il/~vision/DoubleDIP/resources/DoubleDIP.pdf

Project page: http://www.wisdom.weizmann.ac.il/~vision/DoubleDIP/


sketch

If you find our work useful in your research or publication, please cite it:

@article{DoubleDIP,
author = {Gandelsman, Yossi and Shocher, Assaf and Irani, Michal},
year = {2019},
month = {6},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
title = {"Double-DIP": Unsupervised Image Decomposition via Coupled Deep-Image-Priors}
}

Further comments:

The airlight estimation in the dehazing part of the code uses the code provided by "Blind Dehazing Using Internal Patch Recurrence".

The saliency detection that is used for segmentation hints provided by Context-Aware Saliency Detection, by Gofman et al. After applying this saliency detection, we thresholded it using bg_fg_prep.py.

The code is provided as-is for academic use only and without any guarantees. Please contact the author to report any bugs.