Extract faces from photos

Automatic face detection and portrait extraction

Face detection identifies the location and boundaries of human faces within an image. This tool uses the browser's built-in FaceDetector API to find every face in a photo, then crops each one into a separate portrait image. The result is a collection of individual headshots extracted from group photos, event shots, or any image containing people.

The FaceDetector API uses machine learning models built into the browser engine to locate faces regardless of angle, lighting, or partial occlusion. It returns bounding boxes for each detected face, which the tool then expands with configurable padding to include hair, neck, and shoulders for natural-looking portrait crops.

For browsers that do not support the FaceDetector API (Firefox, some Safari versions), the tool provides a manual fallback where you can draw rectangles around faces yourself. This ensures the workflow is available everywhere, even if automatic detection is not.

Face extraction is particularly valuable for event photographers who need to deliver individual portraits from group shots, HR departments processing team photos for directory listings, and social media managers who need profile-sized crops from larger images without manual editing of each face.

Common use cases

• Event photography delivery: Extract individual portraits from group photos at conferences, weddings, or corporate events to deliver personalized headshots to each attendee.
• Employee directory photos: Process team photos into individual headshots for company directories, org charts, and internal communication platforms without manual cropping.
• Social media profile pictures: Quickly extract face crops from full-body or group photos to create properly framed profile pictures for social media accounts.
• Photo organization and tagging: Extract faces from a photo library to create a visual index of people, making it easier to sort and tag large collections.

Technical details

Detection API

Uses the browser's FaceDetector API (Shape Detection API). This runs on-device machine learning models without sending images to any server.

Crop padding

Configurable padding around each detected face bounding box. Default padding includes forehead, chin, and some shoulder area for natural portrait framing.

Browser support

FaceDetector is available in Chrome and Edge on desktop and Android. Safari and Firefox do not yet implement it. Manual rectangle drawing works everywhere.

Output

Each detected face exports as a separate PNG file. Multiple faces from one image download as a ZIP archive with sequential numbering.

How to extract faces from a photo online

Detect and crop every face in a photo automatically.

1. Drop a photo into the face extractor.
2. Wait for the FaceDetector API to find all faces.
3. Adjust the crop padding around each detected face.
4. Download individual portraits or the entire set as a ZIP.

Frequently asked questions