GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → Ding-Liu → NLRN

Ding-Liu / NLRN

Licence: MIT license
Code for Non-Local Recurrent Network for Image Restoration (NeurIPS 2018)

Programming Languages

python
139335 projects - #7 most used programming language

Labels

image-denoisingimage-super-resolution

Projects that are alternatives of or similar to NLRN

Awesome-low-level-vision-resources
A curated list of resources for Low-level Vision Tasks
Stars: ✭ 35 (-78.79%)
Mutual labels:  image-denoising, image-super-resolution
SwinIR
SwinIR: Image Restoration Using Swin Transformer (official repository)
Stars: ✭ 1,260 (+663.64%)
Mutual labels:  image-denoising, image-super-resolution
IDN-pytorch
paper implement : Fast and Accurate Single Image Super-Resolution via Information Distillation Network
Stars: ✭ 40 (-75.76%)
Mutual labels:  image-super-resolution
Uformer
[CVPR 2022] Official repository for the paper "Uformer: A General U-Shaped Transformer for Image Restoration".
Stars: ✭ 415 (+151.52%)
Mutual labels:  image-denoising
RDN-pytorch
PyTorch implementation of Residual Dense Network for Image Super-Resolution (CVPR 2018)
Stars: ✭ 91 (-44.85%)
Mutual labels:  image-super-resolution
strollr2d icassp2017
Image Denoising Codes using STROLLR learning, the Matlab implementation of the paper in ICASSP2017
Stars: ✭ 22 (-86.67%)
Mutual labels:  image-denoising
R2Net
Pytorch code for "Attention Based Real Image Restoration", IEEE Transactions on Neural Networks and Learning Systems, 2021
Stars: ✭ 19 (-88.48%)
Mutual labels:  image-denoising
DudeNet
Designing and Training of A Dual CNN for Image Denoising (Knowledge-based Systems, 2021)
Stars: ✭ 45 (-72.73%)
Mutual labels:  image-denoising
PESR
Official code (Pytorch) for paper Perception-Enhanced Single Image Super-Resolution via Relativistic Generative Networks
Stars: ✭ 28 (-83.03%)
Mutual labels:  image-super-resolution
Reproducible Image Denoising State Of The Art
Collection of popular and reproducible image denoising works.
Stars: ✭ 1,776 (+976.36%)
Mutual labels:  image-denoising
Awesome-ICCV2021-Low-Level-Vision
A Collection of Papers and Codes for ICCV2021 Low Level Vision and Image Generation
Stars: ✭ 163 (-1.21%)
Mutual labels:  image-denoising
PRIDNet
Code for the paper "Pyramid Real Image Denoising Network"
Stars: ✭ 47 (-71.52%)
Mutual labels:  image-denoising
Image-Denoising-with-Deep-CNNs
Use deep Convolutional Neural Networks (CNNs) with PyTorch, including investigating DnCNN and U-net architectures
Stars: ✭ 54 (-67.27%)
Mutual labels:  image-denoising
sparselandtools
✨ A Python package for sparse representations and dictionary learning, including matching pursuit, K-SVD and applications.
Stars: ✭ 55 (-66.67%)
Mutual labels:  image-denoising
Image Super Resolution
🔎 Super-scale your images and run experiments with Residual Dense and Adversarial Networks.
Stars: ✭ 3,293 (+1895.76%)
Mutual labels:  image-super-resolution
SRCNN-pytorch
PyTorch implementation of Image Super-Resolution Using Deep Convolutional Networks (ECCV 2014)
Stars: ✭ 176 (+6.67%)
Mutual labels:  image-super-resolution
FSRCNN-pytorch
PyTorch implementation of Accelerating the Super-Resolution Convolutional Neural Network (ECCV 2016)
Stars: ✭ 115 (-30.3%)
Mutual labels:  image-super-resolution
HCFlow
Official PyTorch code for Hierarchical Conditional Flow: A Unified Framework for Image Super-Resolution and Image Rescaling (HCFlow, ICCV2021)
Stars: ✭ 140 (-15.15%)
Mutual labels:  image-super-resolution

Non-Local Recurrent Network for Image Restoration (NeurIPS 2018)

Paper | Bibtex

WIP: fast evaluation with custom ops

An older version of the NLRN code can be found here.

Usage

Denoising

Preparing BSD500 for training

mkdir -p data/bsd500
wget -O data/bsd500/BSR_bsds500.tgz http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/BSR/BSR_bsds500.tgz
`cd data/bsd500 && tar -xvf BSR_bsds500.tgz`
mkdir -p data/bsd500/flist1
find data/bsd500/BSR/BSDS500/data/images/train/*.jpg data/bsd500/BSR/BSDS500/data/images/test/*.jpg > data/bsd500/flist1/train.flist
find data/bsd500/BSR/BSDS500/data/images/val/*.jpg > data/bsd500/flist1/eval.flist

Preparing Set12 and BSD68 for evaluation

git clone https://github.com/cszn/DnCNN.git data/denoise
find data/denoise/testsets/Set12/*.png > data/set12.flist
find data/denoise/testsets/BSD68/*.png > data/bsd68.flist

Training on flist1 (train and test) of BSD500

python trainer.py --dataset denoise --train-flist data/bsd500/flist1/train.flist --eval-flist data/bsd500/flist1/eval.flist --model nlrn --job-dir debug
# or incremental trainer by number of recurrent states
python incremental_trainer.py --dataset denoise --train-flist data/bsd500/flist1/train.flist --eval-flist data/bsd500/flist1/eval.flist --model nlrn --job-dir debug

Pre-trained models

12 recurrent states/with correlation propagation: sigma 15, sigma 25, sigma 50.

15 recurrent states/without correlation propagation: sigma 15, sigma 25, sigma 50.

Prediction on Set12 and BSD68

python -m datasets.denoise --noise-sigma SIGMA --model-dir MODEL_DIR --input-dir data/denoise/testsets/Set12 --output-dir ./output/Set12
python -m datasets.denoise --noise-sigma SIGMA --model-dir MODEL_DIR --input-dir data/denoise/testsets/BSD68 --output-dir ./output/BSD68

MODEL_DIR is the directory of tf.saved_model and located in export/Servo/ of job_dir.

Super-resolution

Preparing Set5 Set14 BSD100 Urban100 for evaluation

wget -O data/SR_testing_datasets.zip http://vllab.ucmerced.edu/wlai24/LapSRN/results/SR_testing_datasets.zip
`cd data/ && unzip SR_testing_datasets.zip`

Bibtex

@inproceedings{liu2018non,
  title={Non-Local Recurrent Network for Image Restoration},
  author={Liu, Ding and Wen, Bihan and Fan, Yuchen and Loy, Chen Change and Huang, Thomas S},
  booktitle={Advances in Neural Information Processing Systems},
  pages={1680--1689},
  year={2018}
}
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].