Face Verification — Simple Face Recognition for Flutter #

Add secure face recognition to your Flutter app in minutes. This plugin lets you register users' faces and verify their identity later — all processing happens on-device for privacy and works offline.

Perfect for attendance systems, secure login, user verification, and access control apps.

🚀 What You Can Do #

• Register faces: Take a photo and save someone's face profile
• Verify identity: Check if a new photo matches any registered person
• Check registration status: Verify if someone is already registered
• Works offline: No internet required, all processing on-device
• Privacy-first: Face data never leaves the device
• Simple API: Just a few main functions to get started

📱 Quick Demo #

// 1. Initialize (do this once when your app starts)
await FaceVerification.instance.init();

// 2. Register someone's face
await FaceVerification.instance.registerFromImagePath(
  id: 'john_doe',
  imagePath: '/path/to/johns_photo.jpg',
  imageId: 'profile_pic',
);

// 3. Later, verify if a new photo matches John
final matchId = await FaceVerification.instance.verifyFromImagePath(
  imagePath: '/path/to/new_photo.jpg',
  threshold: 0.70, // How strict the match should be (0-1)
);

if (matchId == 'john_doe') {
  print('Welcome back, John!');
} else {
  print('Face not recognized');
}

📦 Installation #

Add to your pubspec.yaml:

dependencies:
  face_verification: ^0.0.6

Run:

flutter pub get

That's it! The face recognition model is included automatically.

🛠️ Platform Setup #

iOS Requirements #

• iOS 15.5 or newer
• Xcode 15.3.0+

Add this to your ios/Podfile:

platform :ios, '15.5'

post_install do |installer|
  installer.pods_project.build_configurations.each do |config|
    config.build_settings["EXCLUDED_ARCHS[sdk=*]"] = "armv7"
    config.build_settings['IPHONEOS_DEPLOYMENT_TARGET'] = '15.5'
  end
  
  installer.pods_project.targets.each do |target|
    flutter_additional_ios_build_settings(target)
    target.build_configurations.each do |config|
      config.build_settings['IPHONEOS_DEPLOYMENT_TARGET'] = '15.5'
    end
  end
end

Android Requirements #

Update your android/app/build.gradle:

android {
    compileSdkVersion 35
    
    defaultConfig {
        minSdkVersion 21
        targetSdkVersion 35
    }
}

🎯 Complete Usage Guide #

Step 1: Initialize the Plugin #

Do this once when your app starts, typically in main() or your first screen:

import 'package:face_verification/face_verification.dart';

class MyApp extends StatefulWidget {
  @override
  _MyAppState createState() => _MyAppState();
}

class _MyAppState extends State<MyApp> {
  @override
  void initState() {
    super.initState();
    _initializeFaceVerification();
  }
  
  Future<void> _initializeFaceVerification() async {
    try {
      await FaceVerification.instance.init();
      print('Face verification ready!');
    } catch (e) {
      print('Failed to initialize: $e');
    }
  }
}

Step 2: Register a Person's Face #

When someone needs to register (like employee onboarding or user signup):

Future<void> registerNewPerson(String personId, String photoPath) async {
  try {
    final result = await FaceVerification.instance.registerFromImagePath(
      id: personId,           // Unique ID like 'employee_123' or 'user_john'
      imagePath: photoPath,   // Path to their photo
      imageId: 'main_photo',  // Required: label for this photo
    );
    
    print('✅ Successfully registered: $result');
    // Show success message to user
    
  } catch (e) {
    print('❌ Registration failed: $e');
    // Handle errors (no face detected, multiple faces, etc.)
  }
}

Step 3: Verify Someone's Identity #

When someone tries to log in or access something:

Future<void> verifyPerson(String photoPath) async {
  try {
    final matchId = await FaceVerification.instance.verifyFromImagePath(
      imagePath: photoPath,
      threshold: 0.70,  // 0.70 = good balance, 0.80 = stricter, 0.60 = more lenient
    );
    
    if (matchId != null) {
      print('✅ Welcome back, $matchId!');
      // Grant access or log them in
    } else {
      print('❌ Face not recognized');
      // Show error or fallback to password
    }
    
  } catch (e) {
    print('❌ Verification failed: $e');
  }
}

Step 4: Check Specific Person (Optional) #

If you want to verify against one specific person instead of everyone:

Future<void> verifySpecificPerson(String photoPath, String expectedPersonId) async {
  final matchId = await FaceVerification.instance.verifyFromImagePath(
    imagePath: photoPath,
    threshold: 0.70,
    staffId: expectedPersonId,  // Only check against this person
  );
  
  if (matchId == expectedPersonId) {
    print('✅ Identity confirmed for $expectedPersonId');
  } else {
    print('❌ Not a match for $expectedPersonId');
  }
}

🔧 Registration Status & Management Functions #

Check Registration Status #

Before registering or to prevent duplicates:

// Check if someone is already registered by ID
final isRegistered = await FaceVerification.instance.isFaceRegistered('john_doe');
if (isRegistered) {
  print('User john_doe is already registered');
} else {
  print('User john_doe is not registered yet');
}

// Check if specific ID + imageId combination exists
final hasSpecificImage = await FaceVerification.instance.isFaceRegisteredWithImageId(
  'john_doe', 
  'profile_pic'
);
if (hasSpecificImage) {
  print('Exact match found - same person with same image');
}

Use cases for registration checks:

• Prevent duplicate registrations
• Show appropriate UI (register vs re-register)
• Validate before allowing re-registration
• Check if user needs to register first

Other Management Functions #

// List all registered people
final records = await FaceVerification.instance.listRegisteredAsync();
print('${records.length} people registered');

// Remove someone's registration
await FaceVerification.instance.deleteRecord('employee_123');

// Clean up when app closes (optional)
await FaceVerification.instance.dispose();

⚙️ Understanding the Threshold #

The threshold parameter controls how strict face matching is:

• 0.60: More lenient (might accept similar-looking people)
• 0.70: Balanced (recommended for most apps)
• 0.80: Strict (reduces false positives, might reject valid matches)
• 0.90: Very strict (use for high-security applications)

📸 Tips for Best Results #

Good photos:

• Clear, front-facing face
• Good lighting
• No sunglasses or masks
• Single person in frame

Avoid:

• Blurry images
• Multiple faces
• Very dark photos
• Extreme angles

🚨 Common Issues & Solutions #

"No face detected"

• Make sure photo has a clear, visible face
• Check lighting and image quality

"Multiple faces detected"

• Crop photo to show only one person
• Use photos with single subjects

"ID already exists" error

• Check if user is already registered using isFaceRegistered()
• Use different imageId for re-registration
• Delete existing record first if needed

iOS build errors

• Make sure you excluded armv7 architecture as shown above
• Test on physical device, not simulator

Low accuracy

• Adjust threshold value
• Use higher quality photos
• Re-register with better photos

🎨 Custom Model (Advanced) #

Want to use your own face recognition model?

await FaceVerification.instance.init(
  modelAsset: 'assets/models/my_custom_model.tflite',
  numThreads: 4,
);

Add the model to your pubspec.yaml:

flutter:
  assets:
    - assets/models/my_custom_model.tflite

📱 Example App #

Check out the complete working example in the example/ folder:

cd example
flutter run

🆘 Need Help? #

Having issues? Please open an issue with:

• Your platform (iOS/Android) and versions
• Device model
• Error messages
• Sample code that isn't working

📄 License #

MIT License - see LICENSE file for details.

⭐ Found this helpful? Please star the repo and leave a review on pub.flutter-io.cn!