This project is a web application for upscaling images using the ESRGAN (Enhanced Super-Resolution Generative Adversarial Networks) model. ESRGAN is a powerful model for image super-resolution, enabling high-quality upscaling of images.
- High-quality image upscaling using ESRGAN.
- Supports both GPU (for faster processing) and CPU (if GPU is not available).
To run this project, you will need to install the following dependencies:
- CUDA (for GPU acceleration)
- PyTorch
- glob2
- OpenCV
-
Install CUDA: Follow the instructions on the NVIDIA website to install CUDA.
-
Install PyTorch: You can install PyTorch with CUDA support by following the instructions on the PyTorch website.
-
Install Python packages:
pip install glob2 opencv-python
Before running the application, ensure the ESRGAN model file (RRDB_ESRGAN_x4.pth) is downloaded from the provided Google Drive link and placed in the models folder.
Download the ESRGAN model file from Google Drive.
This is an intensive program that benefits significantly from GPU acceleration. If a GPU is not available, the application will use the CPU for processing, which will be slower.
This project incorporates code and models from the ESRGAN project by Xintao Wang, licensed under the Apache License 2.0.
This project is licensed under the Apache License 2.0.