expo-blur-face

On-device face detection and blurring for React Native / Expo.

Uses Vision on iOS and ML Kit on Android — no server calls, no FFmpeg, no permissions required. Just install and blur.

Features

• On-device — 100% private, works fully offline
• Zero config — no config plugin, no npx expo prebuild changes required
• Fast — hardware-accelerated detection via native frameworks
• Lightweight — iOS uses built-in Vision + Core Image; Android adds only ML Kit
• Customizable — control blur intensity, padding, and output format (JPG / PNG)
• Multi-face — detects and blurs every face in the image simultaneously

Installation

npm install expo-blur-face
# or
yarn add expo-blur-face

No config plugin needed. Works with bare and managed Expo workflows out of the box.

Usage

import { blurFaces } from 'expo-blur-face';

const result = await blurFaces(imageUri, {
  blurIntensity: 30,  // 1-100 (default: 30)
  padding: 10,        // extra px around face bounds (default: 10)
  output: 'jpg',      // 'jpg' | 'png' (default: 'jpg')
  quality: 0.9,       // JPEG quality 0.0-1.0 (default: 0.9)
});

console.log(result.uri);           // file:///path/to/blurred.jpg
console.log(result.facesDetected); // number of faces found and blurred

If no faces are detected, result.uri is the original URI and facesDetected is 0 — no copy is made.

With image picker

import * as ImagePicker from 'expo-image-picker';
import { blurFaces } from 'expo-blur-face';

async function pickAndBlur() {
  const picked = await ImagePicker.launchImageLibraryAsync({
    mediaTypes: ['images'],
    quality: 1,
  });

  if (picked.canceled) return;

  const result = await blurFaces(picked.assets[0].uri, {
    blurIntensity: 40,
    padding: 15,
  });

  console.log(`Blurred ${result.facesDetected} face(s) -> ${result.uri}`);
}

API

blurFaces(fileUrl, options?)

Detects all faces in an image and applies a blur to each detected face region.

Parameters

| Parameter | Type | Required | Description | |-----------|------|----------|-------------| | fileUrl | string | Yes | file:// URI of the image to process | | options | BlurFacesOptions | No | Blur configuration (see below) |

BlurFacesOptions

| Option | Type | Default | Description | |--------|------|---------|-------------| | blurIntensity | number | 30 | Blur strength — 1 (subtle) to 100 (heavy) | | padding | number | 10 | Pixels to expand beyond the detected face bounding box | | output | 'jpg' \| 'png' | 'jpg' | Output image format | | quality | number | 0.9 | JPEG compression quality 0.0-1.0, ignored for PNG |

Returns: Promise<BlurFacesResult>

| Field | Type | Description | |-------|------|-------------| | uri | string | file:// URI of the output image | | facesDetected | number | Number of faces detected and blurred |

How It Works

iOS: VNDetectFaceRectanglesRequest (Vision framework) finds face bounding boxes, then CIGaussianBlur + CISourceAtopCompositing (Core Image) composites blurred regions back onto the original. Output is written to NSTemporaryDirectory() with EXIF preserved.

Android: Google ML Kit Face Detection returns face bounding boxes, then a StackBlur algorithm is applied directly to each face region on the Bitmap. Output is written to context.cacheDir with EXIF preserved.

Requirements

| | Minimum | |---|---| | Expo SDK | 51+ | | iOS | 15.1+ | | Android | API 24 (Android 7.0)+ | | React Native | 0.73+ |

Changelog

1.0.0 — 2026-03-12

• Initial release
• On-device face blurring via Vision (iOS) and ML Kit (Android)
• Configurable blur intensity, padding, output format
• EXIF data preservation
• Multi-face support

Contributing

Pull requests are welcome. For major changes, please open an issue first.

git clone https://github.com/faeizfurqan17/expo-blur-face.git
cd expo-blur-face
npm install
cd example && npm install
npx expo run:ios      # or run:android

License

MIT © Faeiz Furqan