Blurring faces in videos: a privacy guide for creators

Every vlog, protest clip, school event and dash cam video contains people who didn't choose to be on the internet. Blurring them is often courteous, sometimes legally wise and occasionally essential for someone's safety. Here's how to decide and how to do it properly.

Updated 1 October 20269 min readBeginner

In short: blur children, bystanders in sensitive places, and anyone who could be harmed by being identified. Turn on Blur faces, then scrub every second — automatic detection misses small, turned and shadowed faces. Check number plates, badges and spoken names too.

This guide is general information, not legal advice. Privacy and image-rights laws differ between countries; if you're unsure about a specific video, ask a qualified adviser.

When to blur

In many countries you may film in public places, but "allowed" and "responsible" aren't the same thing. Consider blurring when:

  • Children appear who aren't yours, or whose parents haven't agreed. Schools, sports clubs and parks are the common cases.
  • The place is sensitive — hospitals, clinics, places of worship, shelters, courts, protests.
  • Someone is shown in a bad light — an argument, an accident, a moment of embarrassment. Your video could follow them for years.
  • Your video is commercial. Using a recognisable person in advertising usually needs their permission.
  • Someone asks. Honour requests to remove or blur people quickly — it's decent, and platforms often require it.

In the EU and UK, data-protection law (GDPR) treats a recognisable face as personal data, which is one reason many European creators blur by default.

Beyond faces

People are identified by much more than their faces:

  • Number plates, house numbers and street signs near someone's home.
  • Name badges, lanyards, school or work uniforms with logos.
  • Distinctive tattoos, clothing or a wheelchair in a small community.
  • Screens and documents — phone notifications, letters, ID cards.
  • Reflections in windows, mirrors and sunglasses.
  • Voices and names said aloud.

How automatic blurring works

VDOAPP uses a face-detection model (MediaPipe BlazeFace) on every frame as the video plays and exports. It returns a box around each face it finds; the editor enlarges each box to take in hair and chin, then pixelates it inside an oval about nine blocks across. Because it runs frame by frame, it follows faces as they move — no manual tracking.

Pixelation, rather than a light blur, is deliberate: mild blurs can sometimes be partly reversed, and heavy mosaic is the treatment viewers instantly recognise as "this person's identity is protected".

Step by step

  1. Add your video and split it so that clips needing blur are separate from ones that don't (your own face on camera, for example).
  2. Select a clip, open the AI tab and tick Blur faces under Faces.
  3. Play the clip and scrub slowly through busy moments.
  4. Cover anything the detector missed with a sticker or a shape overlay, animated with keyframes.
  5. Mute or cut audio where names or identifying details are spoken.

What detectors miss

SituationWhyFallback
Tiny faces far awayToo few pixels to recogniseCover with a blurred PiP rectangle
Side profiles, heads downModel expects facial featuresKeyframed sticker
Deep shadow, backlightLow contrastBrighten with Adjust, then check again
Motion blurFace smeared across framesSplit and cover that moment
Masks, helmets, sunglassesPartly hidden featuresOften detected — verify

Why local processing matters here

Many online blurring services upload your footage to a server to process it. That means the very faces you want to protect pass through someone else's computer. In VDOAPP the detection model runs in your browser, so unblurred footage never leaves your device. For journalists, activists, teachers and anyone handling sensitive footage, that isn't a detail — it's the point.

Pre-publish checklist

  1. Watch the full export at normal speed, then scrub the busy parts.
  2. Check the first and last frames and every transition — faces can appear for a split second.
  3. Check reflections, screens and documents.
  4. Listen for names, addresses and phone numbers.
  5. Check your thumbnail — it's easy to blur the video and forget the cover image.

Frequently asked questions

Can blurred faces be unblurred?

Heavy pixelation destroys the detail needed to reconstruct a face. That's why the tool uses a coarse mosaic rather than a light blur.

Can I blur everyone except me?

Blur faces works per clip on every detected face. Keep your own talking shots as separate clips without it, or cover specific other people with keyframed stickers.

Does it work on photos?

Yes — select a photo clip and tick Blur faces in the same way.