A LITTLE PRIVACY. A LOT OF PERSONALITY.
Cover little faces with an emoji or a character made from your photo.
Add your photo. We’ll suggest faces to check.You choose who gets covered.
STEP 3 · DOWNLOAD
These are flattened copies with neutral filenames. Original GPS and camera metadata are not copied. Check clothing, backgrounds and other identifying details before sharing.
Your originals stay untouched.
YOU MAKE THE MATCH
Looking for this face in your other photos. Select only the faces you want covered. Similarity can be wrong, especially across ages, angles and lighting.
Experimental matching runs here. Face comparisons are discarded when you close this panel. Photos still need your final check.
SMALL COVERS. CLEAR BOUNDARIES.
Your photos are processed in this browser.Face detection and optional face matching run on your device. Photo analysis stays in this browser; the editor does not upload photos to an AI service and has no analytics. A local age model helps sort suggestions, which can mistake children and adults. You make the choice. Photo characters use measured outlines, features and colours; hair is approximate and the illustration may retain identifying traits. The hosting provider receives ordinary website requests, such as your IP address and browser information.
You choose who to cover.Suggestions can miss faces, especially in side views or busy photos. Large photos get overlapping scans to look for smaller faces. Suggestions marked “check” have lower detector confidence. Add a cover manually whenever you need one. Check every result.
Fresh files, no hidden original layer.Downloads contain new JPEGs with neutral names. Covers have a solid backing; transparent stickers cannot create holes in the covered area.
A face cover is not guaranteed anonymity.Clothing, locations, other people and visible text may still identify someone. Video lab processes short clips locally. Tracking can drift or miss people; choose one target and review the full result. Unresolved frames need a correction or an opaque whole-frame cover. Exports are silent and timing may vary.
JPG, PNG, WebP & HEIC. Photos larger than 2,560 pixels on their longest side are resized for editing and export. The batch is limited to 40 photos and 48 megapixels of working images. HEIC files are converted on your device; only the first image is used from multi-image files. Closing or clearing this tab discards the editing session. Saved emoji and avatar recipes stay in browser storage until you choose “Forget saved looks”; photos, face comparisons and uploaded stickers are not saved there.
VIDEO LAB · LOCAL PROTOTYPE
One chosen face · clips up to 15 seconds · 960px · 24 output frames/s · silent export
Choose a short clip. It stays in this browser.
Use ← / → to step through frames; hold Shift to jump one second. Slider keys adjust the focused slider.
Tap the target face, or tap anywhere to add a cover. Drag it to correct. Tracking stops at a missing or ambiguous match.
A correction sets an anchor on this frame. Track forward again to continue through to the next anchor.
Mark the first and last frames to cover.
Suggestions cannot tell you that a clip is anonymous. Check other faces, clothing, reflections and background details. Unresolved frames appear fully covered during the masked preview. Original audio and metadata are not copied to the new recording.