01 / Drop
Choose a video
A street clip, an event video or a recording with people in the background.
Free tool
Hide the faces of people in a video before you post it: passers-by in a street clip, people in the background of a recording, or the audience at an event. Every face is found and followed frame by frame, then pixelated, blurred or blacked out. A small AI face finder runs in this tab, so the video is never uploaded.
This runs in your browser. Your videos are never uploaded.
How it works
01 / Drop
A street clip, an event video or a recording with people in the background.
02 / Pick
Pixelate, blur or black boxes. Tick small faces if people are far from the camera.
03 / Check
The faces are followed through the video. Watch the result before you post it.
Questions
No. The face finder is downloaded to your browser and every frame is checked there, so the video never leaves your computer.
YuNet, a small open face detection model from the OpenCV project under the MIT licence, looks at every frame. Each box follows its face smoothly and stays on for a third of a second after the face is lost, so a face that turns away for a moment stays hidden.
Not always. Faces seen from the side or behind, very small or blurry faces, and faces that are partly covered can be missed for a few frames. Watch the whole result before you share it, especially when hiding someone matters.
Not in a video yet: it hides every face it finds. For a photo, the blur faces in a photo tool lets you remove or add boxes.
Yes. The picture is re-encoded as H.264 because every frame changes; the sound is kept.