Image Face Upscale Restoration by malekradwan130

Image Face Upscale Restoration by malekradwan130

Upscales images and restores faces in them, for people who want sharper pictures of people. Its title lists GFPGAN, RestoreFormerPlusPlus and CodeFormer.

Made by malekradwan130 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 Space is a Gradio page for enlarging images and restoring the faces in them. Its title lists three face restoration methods: GFPGAN, RestoreFormerPlusPlus and CodeFormer. The metadata also names the models akhaliq/GPEN-BFR-512, CompVis/stable-diffusion and utnah/esrgan. The source does not describe the controls.

Who it suits: People who want a larger, cleaner version of a picture that shows a face.

Industry | Who uses it | How they use it | Example
Photography | Hobby photographers | Enlarge a picture and restore the faces in it | Sharpening a small portrait
Content creation | Designers and editors | Prepare a bigger version of an image with people in it | Making a larger copy of a headshot
Image research | People who test restoration models | Try face restoration in a browser | Running one portrait through the page

If you never work with pictures of people, you can skip it.

Where to use it:

• A web browser, on the Space's page at Hugging Face
• Python on Windows, macOS or Linux, for a copy you run yourself

Use case | Who | What they get back
Upscale a picture | Anyone with a small image | A larger version of the image
Restore a face | People with soft or damaged portraits | The same picture with its face restored
Try restoration models | Developers | A look at how the named models behave

How to set it up: Online, nothing needs installing. For a copy on your own computer you need a few things first.

• Python
• The pip tool
• The packages listed in requirements.txt
• The system packages listed in packages.txt, which Hugging Face installs with apt-get

1. Install the Hugging Face hub package.
2. Download the Space's files into a folder.
3. Move into that folder.
4. Install Gradio at the pinned version.
5. Install the packages from requirements.txt.
6. Start app.py.

pip install -U huggingface_hub
hf download malekradwan130/Image-Upscalece --repo-type space --local-dir Image-Upscalece
cd Image-Upscalece
pip install gradio==5.16.0
pip install -r requirements.txt
python app.py

The source gives no local step for the system packages, so look at packages.txt before you run a copy.

How to use it: Open the Space's page, or start app.py from your own copy, and give it an image to work on. The source documents no controls, so the labels on the page are your guide.

python app.py

Example: You open the page with a small, soft portrait and run it. What you should get back is a larger, sharper picture with the face restored. The source shows no sample output, so judge the result by eye.

For newcomers and experts: Upscaling makes a picture bigger, and face restoration repairs blur and damage on faces. To start, open the page with one photo of a face and look at what comes back.

The title and the metadata point to GFPGAN, RestoreFormerPlusPlus and CodeFormer for faces, with GPEN-BFR-512 and an ESRGAN model among the named models. The source does not say how they are combined. The Space runs on cpu-basic hardware at Hugging Face.

Open the page in a browser
Give it an image
Upscaling and face restoration run
The restored image comes back

Good to know:

• License: Apache 2.0
• Cost to run: none named
• Data it reaches: the images you give it, and the models named in its metadata
• Not for: the source names no forbidden uses

On Hugging Face, 13 November 2026

• Likes: 10
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/malekradwan130/Image-Upscalece
• Files: https://huggingface.co/spaces/malekradwan130/Image-Upscalece/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.