Face Anonymization
Detects the faces in a photo and covers each one with a blurred oval, leaving the rest of the image untouched. One image in, one PNG out at the original size.
Overview
Face Anonymization finds the faces in a photo and covers each one with a blurred oval. The oval is drawn about 20 percent larger than the detected face box. The blur has the same fixed pixel size on every image, so it removes more detail from small faces than from a face that fills much of a high-resolution photo. Pixels outside the ovals are left exactly as they were.
The group sample on this page shows five people standing by the water. All five faces are blurred, while their clothes, hands, the railing and the buildings behind them are untouched.
What you can build
- A privacy pass for street and event photos before they are published
- Preprocessing for image datasets that must not contain identifiable faces
- Internal sharing of workplace or classroom photos with the faces blurred
Inputs
Upload one photo or paste a link to a publicly reachable image. Blur strength and mask shape are fixed.
Outputs
You get one PNG back, and it keeps the input's resolution. In the samples, a 3828 by 2475 group photo and a 512 by 384 portrait came back at exactly those sizes.
Recommended settings
There are no parameters to tune, so put the effort into review: look over each result for faces the detector missed, such as very small faces or faces partly covered by hands or objects.
Limitations
- Only detected faces are blurred. A missed face stays fully visible.
- The oval stops at the face. Hair, ears and the edge of a beard can stay visible, as the bearded portrait sample shows.
- Some shapes still show through the blur. In the group sample, the sunglasses remain faintly visible.
- Clothing, tattoos, the location and other context can still identify a person after the face is hidden.
- One image per run. Video is not supported.
Safety & compliance
The photo you upload still shows every face before the blur, so get consent from the people pictured first.
API quick start
Run face_anonymization with a single API call.
{
"inputImage": "https://your-cdn.com/input.png"
}curl -X POST "https://api.wiro.ai/v1/Run/wiro/face-anonymization" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"inputImage": "https://your-cdn.com/input.png"
}
JSONOther models you might like
Explore similar image generation models.


