luicfrr/react-native-vision-camera-face-detector

What it solves

This library provides a way to integrate real-time face detection into React Native applications. It bridges the gap between the device's camera and face detection algorithms, allowing developers to identify faces, facial landmarks, contours, and expressions in both live camera feeds and static images.

How it works

The project integrates Google MLKit for the underlying face detection logic and works as a plugin for the react-native-vision-camera module. It can be implemented in three ways: using a high-level <Camera /> component for simple setups, using useFaceDetectorOutput for more control over camera outputs, or using useFaceDetector within a frame processor for custom, asynchronous processing of camera frames.

Who it’s for

Mobile app developers using React Native who need to add face detection capabilities—such as filters, identity verification, or interactive UI—to their apps.

Highlights

• Real-time detection: Supports both front and back cameras for live face tracking.
• Detailed facial analysis: Can detect facial landmarks (eyes, nose, mouth), contours, and classifications (e.g., smiling, eyes open).
• Static image support: Works with local files, URIs, or bundled assets without needing the camera.
• Flexible integration: Compatible with Skia Frame Processors for advanced visual overlays.
• Customizable settings: Adjustable performance modes (fast vs. accurate) and minimum face size thresholds.