How to Blur Faces in Photos (For Privacy)
Blur faces in a single photo or a whole batch — automatically detect every face, choose pixelate vs gaussian blur, and keep the file local. No upload required.
Blurring faces before posting a photo isn't paranoia — it's a basic privacy primitive. Tagged photos travel further than you think (re-shares, screenshots, archive sites), and once a face is identifiable in a high-resolution photo on the internet, it's identifiable forever.
Use Dropvert's face blur
Open Blur Faces. Drop your photo. The tool runs face detection in your browser (no upload) and overlays a box around every detected face.
- Click any face's box to toggle whether it gets blurred (useful when you want to blur strangers but keep your own face visible).
- Pick Pixelate for a chunky-pixel look or Gaussian Blur for a smooth blur.
- Adjust the intensity slider — higher = more aggressive blur, harder to reverse-engineer.
- Adjust the expand region slider — increases the blur area beyond just the face box, catching hair, ears, neck.
Click Blur Faces, download.
Manual paint mode
Sometimes the auto-detector misses a face — usually because:
- The face is heavily turned (>60° rotation).
- The face is partially occluded (sunglasses + hat, mask).
- It's a stylized face (cartoon, painting, statue).
For those cases, click + Paint region and drag a rectangle on the preview to add a manual blur area. You can stack multiple manual regions plus auto-detected faces in the same image.
How aggressive should the blur be?
Privacy researchers have shown that mild blur is reversible — algorithms can de-blur low-intensity blur if there's enough context. To make a face robustly anonymized:
- Pixelate at intensity 15-20 (chunky pixels).
- Gaussian at intensity 18-20.
- Set expand region to 15-20% to cover hair and ears.
Sub-15 intensity blur will look anonymized but can be partially reversed by adversarial models.
Batch processing
Got a folder of photos to anonymize? Drop multiple files at once on Blur Faces. The tool processes them all with the same settings and produces a zip download.
For 20+ files, this is much faster than processing individually — face detection takes ~50ms per face, and the per-image overhead dominates over the actual blur cost.
Don't forget metadata
The image's filename, EXIF data (GPS coords, camera serial number, capture timestamp), and any embedded XMP information can identify a person even with a blurred face. Run the output through Metadata Eraser before sharing publicly.
FAQ
Is the photo uploaded anywhere? No. Both face detection (MediaPipe BlazeFace running locally via WebAssembly) and the blur operation run entirely in your browser. The only network traffic is loading the model itself, which is cached after the first use.
Does it work on cartoon / painted faces? The auto-detector is trained on photographic faces and misses stylized ones. Use + Paint region to manually blur cartoon characters, paintings, statues, etc.
What about videos? Blur Faces has a Video mode that runs the same per-frame detection across an MP4 / MOV / WebM. Includes temporal smoothing so the blur stays stable across single-frame detection misses. Output is MP4, capped at 30 seconds.
Tools mentioned in this guide
Related guides
How to Remove EXIF Data From Photos (Phone or Computer)
EXIF data embedded in photos can include GPS coordinates, camera serial numbers, and timestamps. Here's how to strip it before sharing — without uploading anything.
What Is WebP and Should You Use It?
WebP is a modern image format from Google that produces smaller files than JPEG and PNG with comparable quality. Here's when to use it, when not to, and how to convert.
PNG vs JPEG: Which Image Format Should You Use?
PNG is lossless and supports transparency. JPEG is smaller for photographs. The right choice depends on the image content — here's how to decide.