Coarse-to-Fine Embedded PatchMatch and Multi-Scale Dynamic Aggregation for Reference-based Super-Resolution （AAAI2022）
The code framework is mainly modified from BasicSR and MMSR (Now reorganized as MMEditing). Please refer to the original repo for more usage and documents.
[Paper] [Project Page]
- Python >= 3.7
- PyTorch == 1.4
- CUDA 10.0
- GCC 5.4.0
-
Install Dependencies
cd AMSA conda install pytorch=1.4.0 torchvision cudatoolkit=10.0 -c pytorch pip install mmcv==0.4.4 pip install -r requirements.txt -
Install MMSR and DCNv2
python setup.py develop cd mmsr/models/archs/DCNv2 python setup.py build develop
- Train Set: CUFED Dataset
- Test Set: CUFED5 Dataset
Please refer to Datasets.md for pre-processing and more details.
Downloading the pretrained models from GoogleDrive and put them under experiments/pretrained_models folder.
We provide quick test code with the pretrained model.
-
Modify the paths to dataset and pretrained model in the following yaml files for configuration.
./options/test/test_AMSA.yml ./options/test/test_AMSA_mse.yml
-
Run test code for models trained using GAN loss.
python mmsr/test.py -opt "options/test/test_AMSA.yml"Check out the results in
./results. -
Run test code for models trained using only reconstruction loss.
python mmsr/test.py -opt "options/test/test_AMSA_mse.yml"Check out the results in in
./results
Downloading the pretrained feature extraction models from C2-Matching link and put "feature_extraction.pth" under experiments/pretrained_models folder.
All logging files in the training process, e.g., log message, checkpoints, and snapshots, will be saved to ./experiments and ./tb_logger directory.
- Train restoration network.
# add the path to *pretrain_model_feature_extractor* in the following yaml
# the path to *pretrain_model_feature_extractor* is the model obtained in C2-Matching
./options/train/stage3_restoration_gan.yml
python mmsr/train.py -opt "options/train/stage3_restoration_gan.yml"
# if you wish to train the restoration network with only mse loss
# prepare the dataset path and pretrained model path in the following yaml
./options/train/stage3_restoration_mse.yml
python mmsr/train.py -opt "options/train/stage3_restoration_mse.yml"
If you find our repo useful for your research, please consider citing our paper:
@article{xia2022coarse,
title={Coarse-to-Fine Embedded PatchMatch and Multi-Scale Dynamic Aggregation for Reference-based Super-Resolution},
author={Xia, Bin and Tian, Yapeng and Hang, Yucheng and Yang, Wenming and Liao, Qingmin and Zhou, Jie},
booktitle={AAAI},
year={2022}
}