Image Face Upscale Restoration

Image Face Upscale Restoration

Restores and enlarges faces in images with face restoration models such as GFPGAN, RestoreFormer++ and CodeFormer, plus GPEN and ESRGAN. For anyone fixing old or low-quality portraits.

Made by avans06 on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Gradio Space restores and upscales faces in pictures. Its title names three face restoration models, GFPGAN, RestoreFormerPlusPlus and CodeFormer, and its repository path adds GPEN. It also names an ESRGAN upscaler. The README is only the Hugging Face configuration pointer, so the controls are not described there.

Who it suits: It suits people who want to improve a blurry or damaged face photo, such as family archivists and photo hobbyists. It also suits developers who want to compare several face restoration models in one app.

Industry | Who uses it | How they use it | Example
Photography | Photo hobbyists | Sharpen faces in low-quality shots | Upload a soft portrait and compare the models
Archives | Family historians | Bring old scanned portraits back to life | Try a scanned family photo
Design | Retouchers | Get a first pass before manual edits | Run a face through each model
Software | Developers | Compare restoration models side by side | Read app.py and the utils folder

Anyone who needs forensic or evidentiary image work can skip it, since restoration invents plausible detail.

Where to use it: The app runs in three places.

• A web browser, on the page of the Space at Hugging Face
• Windows, macOS or Linux, as a local copy run with Python
• Windows, through the webui.bat file in the repository

Use case | Who | What they get back
Face restoration | Photographers | A cleaner version of a face in a photo
Upscaling | Archivists | A larger image with restored facial detail
Model comparison | Developers | Results from several restoration models on one input

How to set it up: The online page needs nothing installed. A local copy needs the following.

• Python and pip
• Windows, macOS or Linux
• The packages listed in requirements.txt
• The system packages listed in packages.txt, which the Space installs with apt-get

1. Install Python.
2. Install the Hugging Face hub package and download a copy of the Space.
3. Move into the new folder and install Gradio 5.50.0 and the packages in requirements.txt.
4. Start the app with python app.py.

pip install -U huggingface_hub
hf download avans06/Image_Face_Upscale_Restoration-GFPGAN-RestoreFormer-CodeFormer-GPEN --repo-type space --local-dir Image_Face_Upscale_Restoration-GFPGAN-RestoreFormer-CodeFormer-GPEN
cd Image_Face_Upscale_Restoration-GFPGAN-RestoreFormer-CodeFormer-GPEN
pip install gradio==5.50.0
pip install -r requirements.txt
python app.py

How to use it: Open the page, upload an image with a face, choose a restoration model, and run it. The source does not describe the controls, so follow the labels on the page. On a local copy, this command starts the page.

python app.py

Tool | What it does
GFPGAN | A face restoration model named in the title
RestoreFormerPlusPlus | A second face restoration model named in the title
CodeFormer | A third face restoration model named in the title
GPEN-BFR-512 | A GPEN face restoration model the Space names
ESRGAN | An upscaling model the Space names

Example: You upload a small, blurry portrait and choose one of the restoration models. The page returns a sharper, larger version of the image, and you can run the same photo through another model to compare.

For newcomers and experts: If you are new to this, start with one portrait and run it through a single model. Compare the result with the original, then try a second model on the same file.

If you know the field, the Space gathers GAN-based and transformer-based face restoration models behind one interface, with ESRGAN for upscaling the whole image. Each model fills in plausible facial detail rather than recovering true detail, so results differ by input. The page documents no settings, so read app.py and the utils folder for the options.

Upload an image with a face
Pick a restoration model
The face is restored
The image is upscaled and returned

Good to know: Restored faces are estimates, so check results against the original before relying on them. The README holds no usage notes of its own.

• License: Apache 2
• Cost to run: none named
• Data it reaches: the images you upload, and the models the Space names
• Hardware: basic CPU at Hugging Face, so large images may be slow

On Hugging Face, 13 November 2026

• Likes: 179
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/avans06/Image_Face_Upscale_Restoration-GFPGAN-RestoreFormer-CodeFormer-GPEN
• Files: https://huggingface.co/spaces/avans06/Image_Face_Upscale_Restoration-GFPGAN-RestoreFormer-CodeFormer-GPEN/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.