The repository DougHaber/Real-ESRGAN is a fork of ai-forever/Real-ESRGAN which is a fork of xinntao/Real-ESRGAN.
The ai-forever / sberbank-ai version added in an improved Python interface that I wanted to use. It hadn't been updated in a couple of years, and all usage was triggering a couple Torch warnings. The changes in this repository fix those warnings.
These changes have been made available as a pull request to the upstream as well.
As this is unmaintained, usage is not recommended, but if you feel the need, the correct URL is:
pip install git+https://github.com/DougHaber/Real-ESRGAN.gitIn the event that the PR is accepted, this fork will likely be deleted.
All content below this section is unmodified from the upstream repository.
Update 2024-12-07: I made another change to fix a warning with newer versions of huggingface.
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
Real-ESRGAN is an upgraded ESRGAN 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
- Paper (Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data)
- Original implementation
- Huggingface 🤗
pip install git+https://github.com/sberbank-ai/Real-ESRGAN.gitBasic usage:
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')Low quality image:
Real-ESRGAN result:
Low quality image:
Real-ESRGAN result:
Low quality image:
Real-ESRGAN result:
