All Projects → guoyongcs → Drn

guoyongcs / Drn

Licence: mit
Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to Drn

srVAE
VAE with RealNVP prior and Super-Resolution VAE in PyTorch. Code release for https://arxiv.org/abs/2006.05218.
Stars: ✭ 56 (-82.87%)
Mutual labels:  super-resolution
Mirnet
Official repository for "Learning Enriched Features for Real Image Restoration and Enhancement" (ECCV 2020). SOTA results for image denoising, super-resolution, and image enhancement.
Stars: ✭ 247 (-24.46%)
Mutual labels:  super-resolution
Pytorch Srgan
A modern PyTorch implementation of SRGAN
Stars: ✭ 289 (-11.62%)
Mutual labels:  super-resolution
DLSS
Deep Learning Super Sampling with Deep Convolutional Generative Adversarial Networks.
Stars: ✭ 88 (-73.09%)
Mutual labels:  super-resolution
Awesome-ICCV2021-Low-Level-Vision
A Collection of Papers and Codes for ICCV2021 Low Level Vision and Image Generation
Stars: ✭ 163 (-50.15%)
Mutual labels:  super-resolution
Singan
Official pytorch implementation of the paper: "SinGAN: Learning a Generative Model from a Single Natural Image"
Stars: ✭ 2,983 (+812.23%)
Mutual labels:  super-resolution
TMNet
The official pytorch implemention of the CVPR paper "Temporal Modulation Network for Controllable Space-Time Video Super-Resolution".
Stars: ✭ 77 (-76.45%)
Mutual labels:  super-resolution
Pytorch Vdsr
VDSR (CVPR2016) pytorch implementation
Stars: ✭ 313 (-4.28%)
Mutual labels:  super-resolution
GSOC
Repository for Google Summer of Code 2019 https://summerofcode.withgoogle.com/projects/#4662790671826944
Stars: ✭ 61 (-81.35%)
Mutual labels:  super-resolution
Spsr
Pytorch implementation of Structure-Preserving Super Resolution with Gradient Guidance (CVPR 2020)
Stars: ✭ 280 (-14.37%)
Mutual labels:  super-resolution
RCAN-tf
TensorFlow code for ECCV 2018 paper "Image Super-Resolution Using Very Deep Residual Channel Attention Networks"
Stars: ✭ 25 (-92.35%)
Mutual labels:  super-resolution
traiNNer
traiNNer: Deep learning framework for image and video super-resolution, restoration and image-to-image translation, for training and testing.
Stars: ✭ 130 (-60.24%)
Mutual labels:  super-resolution
Caffe Vdsr
A Caffe-based implementation of very deep convolution network for image super-resolution
Stars: ✭ 273 (-16.51%)
Mutual labels:  super-resolution
RAMS
Official TensorFlow code for paper "Multi-Image Super Resolution of Remotely Sensed Images Using Residual Attention Deep Neural Networks".
Stars: ✭ 55 (-83.18%)
Mutual labels:  super-resolution
Sewar
All image quality metrics you need in one package.
Stars: ✭ 299 (-8.56%)
Mutual labels:  super-resolution
Real-ESRGAN-colab
A Real-ESRGAN model trained on a custom dataset
Stars: ✭ 18 (-94.5%)
Mutual labels:  super-resolution
Video2x
A lossless video/GIF/image upscaler achieved with waifu2x, Anime4K, SRMD and RealSR. Started in Hack the Valley 2, 2018.
Stars: ✭ 3,760 (+1049.85%)
Mutual labels:  super-resolution
Edsr Tensorflow
Tensorflow implementation of Enhanced Deep Residual Networks for Single Image Super-Resolution
Stars: ✭ 314 (-3.98%)
Mutual labels:  super-resolution
Toflow
TOFlow: Video Enhancement with Task-Oriented Flow
Stars: ✭ 314 (-3.98%)
Mutual labels:  super-resolution
Tdan Vsr Cvpr 2020
TDAN: Temporally-Deformable Alignment Network for Video Super-Resolution, CVPR 2020
Stars: ✭ 277 (-15.29%)
Mutual labels:  super-resolution

Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution [arXiv]

Pytorch implementation for "Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution".

Dual Regression Scheme

Dependencies

Python>=3.7, PyTorch>=1.1, numpy, skimage, imageio, matplotlib, tqdm

Quickstart (Model Testing)

Results of our pretrained models:

Model Scale #Params (M) PSNR on Set5 (dB)
DRN-S 4 4.8 32.68
8 5.4 27.41
DRN-L 4 9.8 32.74
8 10.0 27.43

You can evaluate our models on several widely used benchmark datasets, including Set5, Set14, B100, Urban100, Manga109. Note that using an old PyTorch version (earlier than 1.1) would yield wrong results.

Please organize the benchmark datasets using the following hierarchy.

- srdata
    - benchmark
         - Set5
            - LR_bicubic
                - X4
                    - babyx4.png

Dual Regression Scheme

You can use the following script to obtain the testing results:

python main.py --data_dir $DATA_DIR$ \
--save $SAVE_DIR$ --data_test $DATA_TEST$ \
--scale $SCALE$ --model $MODEL$ \
--pre_train $PRETRAINED_MODEL$ \
--test_only --save_results
  • DATA_DIR: path to save data
  • SAVE_DIR: path to save experiment results
  • DATA_TEST: the data to be tested, such as Set5, Set14, B100, Urban100, and Manga109
  • SCALE: super resolution scale, such as 4 and 8
  • MODEL: model type, such as DRN-S and DRN-L
  • PRETRAINED_MODEL: path of the pretrained model

For example, you can use the following command to test our DRN-S model for 4x SR.

python main.py --data_dir ~/srdata \
--save ../experiments --data_test Set5 \
--scale 4 --model DRN-S \
--pre_train ../pretrained_models/DRNS4x.pt \
--test_only --save_results

You will obtain the output like this.

An example of model testing.

If you want to load the pretrained dual model, you can add the following option into the command.

--pre_train_dual ../pretrained_models/DRNS4x_dual_model.pt

Training Method

We use DF2K dataset (the combination of DIV2K and Flickr2K datasets) to train DRN-S and DRN-L.

python main.py --data_dir $DATA_DIR$ \
--scale $SCALE$ --model $MODEL$ \
--save $SAVE_DIR$
  • DATA_DIR: path to save data
  • SCALE: super resolution scale, such as 4 and 8
  • MODEL: model type, such as DRN-S and DRN-L
  • SAVE_DIR: path to save experiment results

For example, you can use the following command to train the DRN-S model for 4x SR.

python main.py --data_dir ~/srdata \
--scale 4 --model DRN-S \
--save ../experiments 

Citation

If you use any part of this code in your research, please cite our paper:

@inproceedings{guo2020closed,
  title={Closed-loop Matters: Dual Regression Networks for Single Image Super-Resolution},
  author={Guo, Yong and Chen, Jian and Wang, Jingdong and Chen, Qi and Cao, Jiezhang and Deng, Zeshuai and Xu, Yanwu and Tan, Mingkui},
  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
  year={2020}
}
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].