Blur Face
Automatically detect and blur or pixelate faces in photos, locally.
Runs LocallyNo Upload100% Private
How it works
- 1Select a photo — it is loaded directly into your browser, never uploaded.
- 2The YuNet detector (int8, ~100 KB) runs on ONNX Runtime WASM in a Web Worker.
- 3Each detected face box is covered with pixelation, a Gaussian blur, or a solid fill.
- 4Preview the result and download it as a PNG.
FAQ
Are my photos uploaded to a server?
No. Detection and blurring run entirely inside your browser. Your photo never leaves your device.
Which blurring styles are available?
Pixelate (mosaic), Gaussian blur, and a solid black ellipse. You can switch between them after detection without re-running the model.
What if a face is not detected?
YuNet is strong on frontal and slightly rotated faces. Very small, backlit or profile-only faces may be missed — try a photo where faces are larger than ~60 px.
Is the detection model really only 100 KB?
Yes. YuNet is a tiny single-stage detector from the OpenCV Zoo, quantized to int8 — it loads in well under a second.