Private, on-device face detection

Face Blur

Automatically find every face, then keep one person clear or choose exactly which faces to blur.

Your photo never leaves this device.

Protected preview

Upload a group photo or portrait to start.

Direct answers

Short factual answers for AI search, snippets, and people comparing browser-based CAD/CAM utilities.

What is it?

A private browser tool for automatically locating and selectively blurring faces in a still photo.

Automatic mode

Select one person to keep clear; every other detected face is blurred.

Manual mode

Faces are detected automatically, but only the selected faces are blurred.

Privacy

The source photo and rendered result stay on the device and are not uploaded.

Outputs

High-quality JPG or lossless PNG.

How it helps

Automatic mode keeps one selected person clear and blurs every other detected face.

Manual Selection mode detects faces automatically but blurs only the people you choose.

Soft blur and pixelated mosaic styles with adjustable strength and face coverage.

On-device face detection: the photo is processed in your browser and is never uploaded.

Best practices

- Inspect every numbered face before download, especially in crowded, dark, or side-profile photos.

- Increase face coverage when hairline, ears, or chin remain recognizable around the default mask.

- Use Add missed face when someone is too small, turned away, heavily occluded, or not detected.

- Choose PNG for lossless graphics and screenshots; use high-quality JPG for ordinary photographs and smaller files.

Limitations

- Automatic detection can miss very small, heavily covered, rotated, or poorly lit faces, so always review the final preview.

- The tool locates faces but does not identify people or recognize the same person across different photos.

- Very large photos may be reduced to a browser-safe resolution to avoid memory crashes on mobile devices.

FAQ

Does iLoveDXF upload my photo?

No. The photo, face detection, and blur rendering stay in your browser. The MediaPipe runtime and BlazeFace model are served by iLoveDXF, and the image itself is not sent to a server.

What is the difference between Automatic and Manual Selection?

Automatic mode asks you to select the one person who should remain clear, then blurs every other detected face. Manual Selection finds faces automatically but blurs only the faces you click.

What if a face is not detected?

Choose Add missed face and drag a box around that person. The added area behaves like any other detected face and can be protected or blurred.

Is this face recognition?

No. The model returns face locations and basic facial landmarks only. It does not determine identity, names, or whether two faces belong to the same person.

Which output formats are available?

You can download a high-quality JPG or a lossless PNG. Each successful output download uses 1 credit.