Skip to content
No description, website, or topics provided.
Branch: master
Clone or download
Latest commit 9c7eba8 Sep 23, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Model
ref_utils
README.md
train_multi_warping.py

README.md

Description

This repository contains the pyTorch implementation of ECCV2018 paper ''CrossNet: An End-to-end Reference-based Super Resolution Network using Cross-scale Warping''. Note that we use a batchsize of 4 instead of 2, and a slightly different input for flownet (see usage 1), results with ~0.2dB higher PSNR is achieved.

We would like to thank Yang Tan (https://github.com/tanyang1231) for implementating the code using pyTorch.

prerequisite library

pytorch opencv h5py

usage

  1. prepare an hdf5 file, which contains /img_HR, /img_LR, /img_MDSR, /img_LR_upsample. /img_HR is used as reference input and ground truth, /img_LR is used as low resolution input, /img_MDSR is the MDSR upsampled image, and /img_LR_upsample is bicubically upsampled image. (Different from the original paper, in this version of code, we use Flownet with bicubically upsampled image and reference image to generate optical flow)

  2. mkdir checkpoints_charbonnier_3

  3. python -u train_multi_warping.py --dataset Flower --display 50 --batch_size 2 --step_size 150000 --loss CharbonnierLoss --optim Adam --lr 0.0001 --checkpoints_dir ./checkpoints_charbonnier_3/ --checkpoint 195000 --checkpoint_file ./checkpoints_charbonnier_3/CP195000.pth

citation

Please cite our paper if you find it interesting!

@inproceedings{zheng2018crossnet,
title={CrossNet: An End-to-end Reference-based Super Resolution Network using Cross-scale Warping},
author={Zheng, Haitian and Ji, Mengqi and Wang, Haoqian and Liu, Yebin and Fang, Lu},
booktitle={Proceedings of the European Conference on Computer Vision (ECCV)},
pages={88--104},
year={2018}
}

You can’t perform that action at this time.