Official implementation for Perception-Enhanced Single Image Super-Resolution via Relativistic Generative Networks ECCV Workshop 2018
Please our project if it is helpful for your research
@InProceedings{Vu_2018_ECCV_Workshops},
author = {Vu, Thang and Luu, Tung M. and Yoo, Chang D.},
title = {Perception-Enhanced Image Super-Resolution via Relativistic Generative Adversarial Networks},
booktitle = {The European Conference on Computer Vision (ECCV) Workshops},
month = {September},
year = {2018}
}
- Nvidia GPUs (training takes 1 day on 4 Titan Xp GPUs)
- At least 16G RAM
Python3Pytorch 0.4tensorboardXtqdmimageio
- Train: DIV2K (800 2K-resolution images)
- Valid (for visualization): DIV2K (100 val images), PIRM (100 self-val images)
- Test: Set5, Set14, B100, Urban100, PIRM (100 self-val images), DIV2K (100 val images)
- Download train+val+test datasets
- Download test only dataset
- Download pretrained models including 1 PSNR-optimized model and 1 perception-optimized model
- Download paper results in images of the test datasets
- Download test dataset and put into
data/origin/directory - Download pretrained model and put into
check_pointdirectory - Run
python test.py --dataset <DATASET_NAME> - Results will be saved into
results/directory
- Download train+val+test dataset and put into
data/origin directory - Pretrain with L1 loss:
python train.py --phase pretrain --learning_rate 1e-4 - Finetune on pretrained model with GAN:
python train.py - Models with be saved into
check_point/direcory
- Start tensorboard:
tensorboard --logdir check_point - Enter:
YOUR_IP:6006to your web browser. - Tensorboard when finetuning on pretrained model should be similar to:
- Test perceptual model: follow Quick start
- Interpolate between perceptual model and PSNR model:
python test.py --dataset <DATASET> --alpha <ALPHA>(with alpha being perceptual weight) - Test perceptual quality: refer to PIRM validation code
RED and BLUE indicate best and second best respectively.
