wyannn / ntsdcn

NTSDCN: New Three-Stage Deep Convolutional Image Demosaicking Network

Implementation of "NTSDCN: New Three-Stage Deep Convolutional Image Demosaicking Network "

TCSVT

Installation

The code was developed using Python 3.5 & TensorFlow 1.9.0 & CUDA 9.0.

Code v1.0

Currently, we release our research code for testing. The algorithm was reimplemented by us and got better results than the paper.

Citation

If you use any part of our code, or our method is useful for your research, please consider citing:

@inproceedings{CTDDSCNN2019,
author = {Yan Wang, Shiying Yin, Shuyuan Zhu, Member, IEEE, Zhan Ma, Member, IEEE, Ruiqin Xiong, Member, IEEE, and Bing Zeng, Fellow, IEEE},
title = {NTSDCN: New Three-Stage Deep Convolutional Image Demosaicking Network},
booktitle = {TCSVT}}

Contact

We are glad to hear if you have any suggestions and questions. Please send email to [email protected]

ntsdcn's Issues

testing pretrained weights

Hi,

are the ckpt file providing pretrained results for both Kodak and MCM image sets. I am getting erronous results. Can you give more instruction to test this code?

br,

-Mikko