https://github.com/ai-forever/Real-ESRGAN

PyTorch implementation of Real-ESRGAN model
https://github.com/ai-forever/Real-ESRGAN

esrgan python pytorch real-esrgan real-world-super-resolution super-resolution

Last synced: about 2 months ago
JSON representation

PyTorch implementation of Real-ESRGAN model

• Host: GitHub
• URL: https://github.com/ai-forever/Real-ESRGAN
• Owner: ai-forever
• License: bsd-3-clause
• Created: 2021-11-02T08:41:49.000Z (almost 3 years ago)
• Default Branch: main
• Last Pushed: 2024-04-15T18:19:26.000Z (5 months ago)
• Last Synced: 2024-07-08T11:01:52.302Z (3 months ago)
• Topics: esrgan, python, pytorch, real-esrgan, real-world-super-resolution, super-resolution
• Language: Python
• Homepage:
• Size: 9.93 MB
• Stars: 434
• Watchers: 12
• Forks: 110
• Open Issues: 16
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Real-ESRGAN

PyTorch implementation of a Real-ESRGAN model trained on custom dataset. This model shows better results on faces compared to the original version. It is also easier to integrate this model into your projects.

> This is not an official implementation. We partially use code from the [original repository](https://github.com/xinntao/Real-ESRGAN)

Real-ESRGAN is an upgraded [ESRGAN](https://arxiv.org/abs/1809.00219) trained with pure synthetic data is capable of enhancing details while removing annoying artifacts for common real-world images. 

You can try it in [google colab](https://colab.research.google.com/drive/1YlWt--P9w25JUs8bHBOuf8GcMkx-hocP?usp=sharing) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1YlWt--P9w25JUs8bHBOuf8GcMkx-hocP?usp=sharing)

- [Paper (Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data)](https://arxiv.org/abs/2107.10833)

- [Original implementation](https://github.com/xinntao/Real-ESRGAN)

- [Huggingface 🤗](https://huggingface.co/sberbank-ai/Real-ESRGAN)

### Installation

```bash

pip install git+https://github.com/sberbank-ai/Real-ESRGAN.git

```

### Usage

---

Basic usage:

```python

import torch

from PIL import Image

import numpy as np

from RealESRGAN import RealESRGAN

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

model = RealESRGAN(device, scale=4)

model.load_weights('weights/RealESRGAN_x4.pth', download=True)

path_to_image = 'inputs/lr_image.png'

image = Image.open(path_to_image).convert('RGB')

sr_image = model.predict(image)

sr_image.save('results/sr_image.png')

```

### Examples

---

Low quality image:

![](inputs/lr_image.png)

Real-ESRGAN result:

![](results/sr_image.png)

---

Low quality image:

![](inputs/lr_face.png)

Real-ESRGAN result:

![](results/sr_face.png)

---

Low quality image:

![](inputs/lr_lion.png)

Real-ESRGAN result:

![](results/sr_lion.png)