Real-ESRGAN

PyTorch implemetatio of a Real-ESRGAN model traied o custom dataset. This model shows better results o faces compared to the origial versio. It is also easier to itegrate this model ito your projects.

Real-ESRGAN is a upgraded ESRGAN traied with pure sythetic data is capable of ehacig details while removig aoyig artifacts for commo real-world images.

• Paper
• Origial implemetatio
• Our github

Usage

Code for usig model you ca obtai i our repo.

import torch
from PIL import Image
import umpy as p
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', dowload=True)

path_to_image = 'iputs/lr_image.pg'
image = Image.ope(path_to_image).covert('RGB')

sr_image = model.predict(image)

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

Real-ESRGAN PyTorch implementation of a Real-ESRGAN model trained on custom dataset. This model show