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.

You can try it in google colab

• Paper: Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data
• Official github

Installation

1. Clone repo

   git clone https://https://github.com/sberbank-ai/Real-ESRGAN
   cd Real-ESRGAN

2. Install requirements

   pip install -r requirements.txt

3. Download pretrained weights and put them into weights/ folder

Usage

Basic example:

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')

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')