SRMNet

KO | EN

gitlite — search

#javascript #react #typescript #go #android #linux #docker #awesome #cli #kubernetes #awesome-list #vue #pytorch #open-source

(★ 31)

Selective Residual M-Net for Real Image Denoising

SRMNet 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# [EUSIPCO 2022] [Selective Residual M-Net for Real Image Denoising](https://eurasip.org/Proceedings/Eusipco/Eusipco2022/pdfs/0000469.pdf)   
[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/selective-residual-m-net-for-real-image/image-denoising-on-sidd)](https://paperswithcode.com/sota/image-denoising-on-sidd?p=selective-residual-m-net-for-real-image)

## [Chi-Mao Fan](https://github.com/FanChiMao), Tsung-Jung Liu, Kuan-Hsien Liu   
[![paper](https://img.shields.io/badge/arXiv-Paper-brightgreen)](https://arxiv.org/abs/2203.01645)
[![official_paper](https://img.shields.io/badge/IEEE-Paper-blue)](https://ieeexplore.ieee.org/document/9909521)
[![video](https://img.shields.io/badge/Video-Presentation-F9D371)](https://youtu.be/uju2fSa44h4)
[![slides](https://img.shields.io/badge/Presentation-Slides-B762C1)](https://docs.google.com/presentation/d/1UCFud_-NsZs6pIxFCtFe3IevMjfifO_Q/edit?usp=sharing&ouid=108348190349543369603&rtpof=true&sd=true)
[![Hugging Face Spaces](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Spaces-blue)](https://huggingface.co/spaces/52Hz/SRMNet_real_world_denoising) 
![Visitors](https://api.visitorbadge.io/api/visitors?path=https%3A%2F%2Fgithub.com%2FFanChiMao%2FSRMNet&label=visitors&countColor=%232ccce4&style=plastic)

> Abstract : Image restoration is a low-level vision task which is restoring
the degraded images to noise-free images. With the success
of deep neural networks, especially the convolutional
neural networks suppress the traditional restoration methods
and become the mainstream in the computer vision. To advance
the performance of denoising algorithms, we propose a
blind real image denoising network (SRMNet) by employing
a hierarchical architecture improved from U-Net. We
use a selective kernel with residual block on the hierarchical
structural named M-Net to enrich the multi-scale semantic
information. Furthermore, our SRMNet has competitive
performance results on two synthetic and two realworld
noisy datasets in terms of quantitative metrics and
visual quality.

## Network Architecture 
<table>
 <tr>
 <td colspan="2"><img src = "https://i.imgur.com/GYeypta.png" alt="SRMNet" width="800"> </td> 
 </tr>
 <tr>
 <td colspan="2">Overall Framework of SRMNet</td>
 </tr>
 
 <tr>
 <td> <img src = "https://i.imgur.com/z6Vds87.png" width="400"> </td>
 <td> <img src = "https://i.imgur.com/eaLejBK.png" width="400"> </td>
 </tr>
 <tr>
 <td>Selective Residual Block (SRB)</td>
 <td> Resizing Block (Pixel Shuffle)</td>
 </tr>
</table>

## Quick Run  
You can simply demo on the space of Hugging Face:  
- [**Real denoising**](https://huggingface.co/spaces/52Hz/SRMNet_real_world_denoising)  
- [**AWGN denoising**](https://huggingface.co/spaces/52Hz/SRMNet_AWGN_denoising)

Or test on local environment:

To test the pre-trained models of Denoising on your own images, run
```
python demo.py --input_dir images_folder_path --result_dir save_images_here --weights path_to_models
```
**All pre-trained models can be downloaded at [pretrained_model/README.md](pretrained_model/README.md) or [here](https://github.com/FanChiMao/SRMNet/releases)**

## Train  
To train the restoration models of Denoising. You should check the following components are correct:  
- `training.yaml`:  
  ```
  # Training configuration
  GPU: [0,1,2,3]

VERBOSE: False

MODEL:
    MODE: 'SRMNet_denoise'

# Optimization arguments.
  OPTIM:
    BATCH: 2
    EPOCHS: 100
    # EPOCH_DECAY: [10]
    LR_INITIAL: 1e-4
    LR_MIN: 1e-6
    # BETA1: 0.9

TRAINING:
    VAL_AFTER_EVERY: 1
    RESUME: False
    TRAIN_PS: 256
    VAL_PS: 256
    TRAIN_DIR: 'D:/PycharmProjects/SUNet-main/datasets/Denoising_DIV2K/train'       # path to training data
    VAL_DIR: 'D:/PycharmProjects/SUNet-main/datasets/Denoising_DIV2K/test' # path to validation data
    SAVE_DIR: './checkpoints'           # path to save models and images
  ```
  
- Dataset:  
  The preparation of dataset in more detail, see [Dataset/README.md](Dataset/README.md).  
  
- Train:  
  If the above path and data are all correctly setting, just simply run:  
  ```
  python train.py
  ```  
## Test (Evaluation)

- To run the models of real image denoising, see [test_DND_real_denoising.py](./test_DND_real_denoising.py) and [test_SIDD_real_denoising.py](./test_SIDD_real_denoising.py).  
- To test the PSNR and SSIM of *real image denoising*, see [evaluation_DND.py](./evaluation_DND.py) and [evaluation_SIDD.m](./evaluation_SIDD.m).  
- To test the PSNR and SSIM of *AWGN image denoising*, see the [evaluation.m](./evaluation.m).

## Result 
- AWGN image denoising 
<img src = "https://i.imgur.com/TILnGHa.png" width="800">

- Real image denoising 
<img src = "https://i.imgur.com/vxt6Vs9.png" width="400">

- Comparison of model complexity
<img src = "https://i.imgur.com/Eg0zoWD.png" width="400">

## Visual Comparison

**More visual results can be downloaded at [here](https://github.com/FanChiMao/SRMNet/releases).**

## Citation  
```
@inproceedings{fan2022selective,
  title={Selective Residual M-Net for Real Image Denoising},
  author={Fan, Chi-Mao and Liu, Tsung-Jung and Liu, Kuan-Hsien and Chiu, Ching-Hsiang},
  booktitle={2022 30th European Signal Processing Conference (EUSIPCO)},
  pages={469--473},
  year={2022},
  organization={IEEE}
}
```

## Contact  
If you have any question, feel free to contact qaz5517359@gmail.com

SRMNet

SRMNet 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

SRMNet

SRMNet 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README