A Flutter plugin to use the Firebase ML Kit

August 11, 2021 Firebase, Plugin, plugin

A Flutter plugin to use the Firebase ML Kit.

this is not official package

The flutter team now has the firebase_ml_vision or firebase_ml_custom package for Firebase ML Kit. Please consider trying to use firebase_ml_vision.

Note: This plugin is still under development, and some APIs might not be available yet. Feedback and Pull Requests are most welcome!

Features

Feature	Android	iOS
Recognize text(on device)
Recognize text(cloud)	yet	yet
Detect faces(on device)
Scan barcodes(on device)
Label Images(on device)
Label Images(cloud)	yet	yet
Object detection & tracking	yet	yet
Recognize landmarks(cloud)	yet	yet
Language identification
Translation	yet	yet
Smart Reply	yet	yet
AutoML model inference	yet	yet
Custom model(on device)
Custom model(cloud)

What features are available on device or in the cloud?

Usage

To use this plugin, add mlkit as a dependency in your pubspec.yaml file.

Getting Started

Check out the example directory for a sample app using Firebase Cloud Messaging.

Android Integration

To integrate your plugin into the Android part of your app, follow these steps:

Using the Firebase Console add an Android app to your project: Follow the assistant, download the generated google-services.json file and place it inside android/app. Next, modify the android/build.gradle file and the android/app/build.gradle file to add the Google services plugin as described by the Firebase assistant.

iOS Integration

To integrate your plugin into the iOS part of your app, follow these steps:

Using the Firebase Console add an iOS app to your project: Follow the assistant, download the generated GoogleService-Info.plist file, open ios/Runner.xcworkspace with Xcode, and within Xcode place the file inside ios/Runner. Don’t follow the steps named “Add Firebase SDK” and “Add initialization code” in the Firebase assistant.

Dart/Flutter Integration

From your Dart code, you need to import the plugin and instantiate it:

import 'package:mlkit/mlkit.dart';

FirebaseVisionTextDetector detector = FirebaseVisionTextDetector.instance;

// Detect form file/image by path
var currentLabels = await detector.detectFromPath(_file?.path);

// Detect from binary data of a file/image
var currentLabels = await detector.detectFromBinary(_file?.readAsBytesSync());

custom model interpreter

native sample code

import 'package:mlkit/mlkit.dart';
import 'package:image/image.dart' as img;

FirebaseModelInterpreter interpreter = FirebaseModelInterpreter.instance;
FirebaseModelManager manager = FirebaseModelManager.instance;

//Register Cloud Model
manager.registerRemoteModelSource(
        FirebaseRemoteModelSource(modelName: "mobilenet_v1_224_quant"));

//Register Local Backup
manager.registerLocalModelSource(FirebaseLocalModelSource(modelName: 'mobilenet_v1_224_quant',  assetFilePath: 'ml/mobilenet_v1_224_quant.tflite');


var imageBytes = (await rootBundle.load("assets/mountain.jpg")).buffer;
img.Image image = img.decodeJpg(imageBytes.asUint8List());
image = img.copyResize(image, 224, 224);

//The app will download the remote model. While the remote model is being downloaded, it will use the local model.
var results = await interpreter.run(
        remoteModelName: "mobilenet_v1_224_quant",
        localModelName: "mobilenet_v1_224_quant",
        inputOutputOptions: FirebaseModelInputOutputOptions([
          FirebaseModelIOOption(FirebaseModelDataType.FLOAT32, [1, 224, 224, 3])
        ], [
          FirebaseModelIOOption(FirebaseModelDataType.FLOAT32, [1, 1001])
        ]),
        inputBytes: imageToByteList(image));

// int model
Uint8List imageToByteList(img.Image image) {
    var _inputSize = 224;
    var convertedBytes = new Uint8List(1 * _inputSize * _inputSize * 3);
    var buffer = new ByteData.view(convertedBytes.buffer);
    int pixelIndex = 0;
    for (var i = 0; i < _inputSize; i++) {
      for (var j = 0; j < _inputSize; j++) {
        var pixel = image.getPixel(i, j);
        buffer.setUint8(pixelIndex, (pixel >> 16) & 0xFF);
        pixelIndex++;
        buffer.setUint8(pixelIndex, (pixel >> 8) & 0xFF);
        pixelIndex++;
        buffer.setUint8(pixelIndex, (pixel) & 0xFF);
        pixelIndex++;
      }
    }
    return convertedBytes;
  }

// float model
Uint8List imageToByteList(img.Image image) {
  var _inputSize = 224;
  var convertedBytes = Float32List(1 * _inputSize * _inputSize * 3);
  var buffer = Float32List.view(convertedBytes.buffer);
  int pixelIndex = 0;
  for (var i = 0; i < _inputSize; i++) {
    for (var j = 0; j < _inputSize; j++) {
      var pixel = image.getPixel(i, j);
      buffer[pixelIndex] = ((pixel >> 16) & 0xFF) / 255;
      pixelIndex += 1;
      buffer[pixelIndex] = ((pixel >> 8) & 0xFF) / 255;
      pixelIndex += 1;
      buffer[pixelIndex] = ((pixel) & 0xFF) / 255;
      pixelIndex += 1;
    }
  }
  return convertedBytes.buffer.asUint8List();
}