facecore 0.6.0

Self-contained Flutter plugin for camera-based face liveness, depth, and anti-spoof checks before capturing a photo on Android and iOS.

facecore

A self-contained Flutter plugin that runs a camera-based face liveness flow natively on Android and iOS, then returns a captured JPEG to Dart. The platform side handles camera permission, preview, face detection, motion and depth anti-spoof checks, and image capture — your Dart code just calls one method.

Features

• Native liveness UI on Android (CameraX + ML Kit Face Detection) and iOS (AVFoundation + Vision).
• Configurable challenge actions: blink, turn left, turn right.
• Motion standard-deviation check to reject static photos / replays.
• Optional depth-based anti-spoof check.
• Returns the saved JPEG path plus telemetry (dimensions, motion, depth).

Platform support

Platform	Minimum version
Android	API 21
iOS	13.0

Installation

dependencies:
  facecore: ^0.1.0

Android

Camera permission is declared by the plugin. No additional setup required.

iOS

Add a usage description to your app's ios/Runner/Info.plist:

<key>NSCameraUsageDescription</key>
<string>This app uses the camera to verify your identity.</string>

Usage

import 'package:facecore/facecore.dart';

final facecore = Facecore();

final result = await facecore.captureWithLiveness(
  config: const LivenessConfig(
    actions: [
      LivenessAction.blink,
      LivenessAction.turnLeft,
      LivenessAction.turnRight,
    ],
    requireDepth: true,
    title: 'Verify your identity',
    subtitle: 'Follow the prompts to confirm you are present.',
  ),
);

if (result == null) {
  // User cancelled.
  return;
}

print('Saved to: ${result.imagePath}');
print('Size: ${result.width}x${result.height}');
print('Motion stddev: ${result.motionStdDev}');
print('Depth: ${result.depthMethod.name} score=${result.depthScore}');

captureWithLiveness returns null when the user cancels and throws a FaceCaptureException if the native flow fails (e.g. permission denied, camera unavailable, anti-spoof rejection).

See example/ for a runnable demo.

Configuration

LivenessConfig exposes the knobs that drive the native flow:

Field	Default	Purpose
actions	[blink, turnLeft, turnRight]	Ordered list of liveness challenges to run.
requireDepth	false	Opt-in: verify the face is 3D; rejects photos/flat screens (below).
minFaceFraction	0.18	Minimum face height as a fraction of the frame height.
maxFaceFraction	0.85	Maximum face height (too close).
minMotionStdDev	1.5	Minimum motion variance to accept the capture.
minDepthRange	10.0	Minimum nose relief in mm (TrueDepth check).
minParallaxScore	0.02	Minimum parallax score (motion-based depth check).
holdFrames	6	Frames the user must hold still before capture.
jpegQuality	90	JPEG quality (1–100) for the saved capture.
title	null	Optional title shown on the native screen.
subtitle	null	Optional subtitle shown on the native screen.

Depth check

The depth check is opt-in. When requireDepth is true, the plugin picks the strongest method the device supports and reports it in FaceCaptureResult.depthMethod, with the measurement in depthScore:

Device	Method	How it works	Rejects
iPhone/iPad with Face ID	trueDepth	Reads the TrueDepth depth map; the nose must protrude ≥ minDepthRange mm in front of both the cheeks and the forehead/chin.	Photos (flat or bent), screens showing photos or videos.
Android, iOS without Face ID	parallax	During head turns, fits a planar homography to ~130 facial landmarks. A flat surface fits exactly; a real face cannot. Score is net of detector noise.	Photos and still images on screens, even when tilted/moved.

The parallax method cannot detect a screen replaying a video of a real person, because that video contains genuine 3D motion. Defaults were derived from simulation; log depthScore on your target devices to tune thresholds.

License

MIT — see LICENSE.