3 linesto upscale an image with the high-level API
7K+all-time developer downloads reported for the package
2 modelscombined behind a single call
The problem
Upscaling images usually means assembling two different models, handling their different input formats, and writing glue code before you get a usable result.
Approach
- Combined RealESRGAN for background super-resolution with GFPGAN for facial feature reconstruction.
- Exposed a single high-level API so the common case needs three lines of code.
- Shipped as a standard Python package so it can be installed and used inside existing pipelines.
Stack
- Python
- PyTorch
- OpenCV
- RealESRGAN
- GFPGAN
- Gradio
What these numbers do and do not prove
Download counts are self-reported package figures and change over time. Check the current package registry entry before citing a number.