mediapipe_face_mesh

Bundled files:

• TensorFlow Lite C runtime binaries for Android (arm64-v8a, x86_64) and iOS
• MediaPipe TFLite model
  • face mesh
  • iris
  • short-range face detection
  • full-range dense and sparse face detection

Supported Platforms

• Android(arm64-v8a, x86_64)
• iOS
• Dart SDK: >=3.8.1 <4.0.0

Install

flutter pub add mediapipe_face_mesh

Usage

Create Face Detector Processor

import 'package:mediapipe_face_mesh/mediapipe_face_mesh.dart';

final faceDetectorProcessor = await FaceDetectorProcessor.create(
  model: FaceDetectionModel.fullRange,
  delegate: FaceMeshDelegate.xnnpack,
  maxResults: 1,
  roiScaleY: 1.7,
  roiShiftY: -0.2,
);

FaceDetectionModel selects the bundled detector model: shortRange is the default short-range BlazeFace model, fullRange is the dense full-range model, and fullRangeSparse is the sparse full-range model.

ROI options adjust the detector-produced expandedFaceRect, which is passed to face mesh while keeping the original frame unchanged.

Create Face Mesh Processor

import 'package:mediapipe_face_mesh/mediapipe_face_mesh.dart';

final faceMeshProcessor = await FaceMeshProcessor.create(
  delegate: FaceMeshDelegate.xnnpack,
  enableSmoothing: true,
  enableRoiTracking: true,
  enableIris: true, // default is false; true returns 478 landmarks with 10 iris points
);

When enableIris is enabled, Face Mesh runs an additional iris landmark pass after the base 468-point face mesh result. The final result keeps the existing Face Mesh index layout, updates the eye-region landmarks with more precise eye contour coordinates, and appends 10 iris landmarks at indices 468..477.

Delegate options:

• FaceMeshDelegate.cpu (default)
• FaceMeshDelegate.xnnpack
• FaceMeshDelegate.gpuV2

If the requested delegate is unavailable or fails to initialize, the native runtime falls back to CPU inference.

Input Formats

The package supports two image input types:

• FaceMeshNv21Image Use this for Android camera frames in NV21 layout.
• FaceMeshImage Use this for RGBA or BGRA buffers. This is used for iOS camera frames.

Stream Inference

Use stream inference when processing continuous camera frames. Stream processors take a Stream of frames and return a Stream of results.

final pipeline = FaceMeshInferencePipeline(
  detector: faceDetectorProcessor,
  mesh: faceMeshProcessor,
);
final inferenceStreamProcessor = FaceMeshInferenceStreamProcessor(pipeline);
final frameController = StreamController<FaceMeshNv21Image>();
bool _isBusy = false;

inferenceStreamProcessor
    .processNv21(
      frameController.stream,
      runMeshResolver: (_) => _isMeshActive,
      rotationDegrees: rotationDegrees,
    )
    .listen(_handleInferenceResult, onError: onError);

void _handleInferenceResult(FaceMeshInferenceResult result) {
  _isBusy = false;
  onDetections(result.detectionResult);
  onMeshResult(result.meshResult);
}

void onCameraFrame(FaceMeshNv21Image frame) {
  if (_isBusy) return;
  _isBusy = true;
  frameController.add(frame);
}

Use runMesh: false when an entire stream should run detector-only. Use runMeshResolver when mesh execution should be decided per frame, such as a UI toggle that can change while the stream is active.

For BGRA / RGBA input, use process(...) instead of processNv21(...).

Single Inference

Use single-frame inference in one call without a stream processor.

final pipeline = FaceMeshInferencePipeline(
  detector: faceDetectorProcessor,
  mesh: faceMeshProcessor,
);

final result = pipeline.processNv21(
  nv21Image,
  rotationDegrees: rotationDegrees,
);

final meshResult = result.meshResult;
if (meshResult != null) {
  onResult(meshResult);
}

For detector-only, set runMesh: false.

final result = pipeline.processNv21(
  nv21Image,
  runMesh: false,
  rotationDegrees: rotationDegrees,
);

Close Resource

Explicitly calling close() when the processors are no longer needed is recommended.

faceDetectorProcessor.close();
faceMeshProcessor.close();

Notes

The examples in this README use the v1.6.0+ unified inference API.

FaceMeshInferenceStreamProcessor emits one combined result after detector and mesh inference complete, so detection boxes and mesh landmarks are updated together.

If you need detector boxes to update independently from slower mesh inference, use FaceDetectorStreamProcessor and FaceMeshStreamProcessor separately; see the v1.5.0 README and example app for a two-stage stream pattern.

These separated stream processors are still available in v1.6.0 and later.

Example app

The example app in example/ provides two flows:

A. MediaPipe Face Detector + MediaPipe Face Mesh
B. ML Kit Face Detector + MediaPipe Face Mesh

B depends on the google_mlkit_face_detection package for face detection.

Primary API

• FaceMeshInferencePipeline Runs face detection and face mesh inference in one call for single-frame use.
• FaceMeshInferenceStreamProcessor Wraps FaceMeshInferencePipeline in an async* generator — accepts a Stream of frames and yields a Stream of high-level inference results.
• FaceMeshInferenceResult Contains detector output, selected detection, selected box/ROI, ROI availability, and mesh output.
• FaceDetectorProcessor Runs the bundled MediaPipe short-range, full-range dense, or full-range sparse face detector and returns face boxes, scores, and rotation-aware ROI values such as expandedFaceRect.
• FaceMeshProcessor Runs face mesh inference and returns normalized 3D landmarks, mesh triangles, the detected face rect, score, and input image size.
• FaceMeshNv21Image Input wrapper for Android NV21 camera frames.
• FaceMeshImage Input wrapper for RGBA or BGRA pixel buffers.
• NormalizedRect Rotation-aware normalized ROI used to restrict face mesh inference.
• FaceMeshBox Pixel-space bounding box that can be converted into an ROI internally.
• FaceMeshResult Result object containing landmarks, triangles, rect, score, imageWidth, and imageHeight. Face mesh returns 468 landmarks by default, or 478 landmarks when enableIris is enabled. Pixel-space helpers such as landmarkAsOffset(...) and landmarksAsOffsets(...) support rotation and horizontal mirror mapping for preview overlays.