Flutter Firebase 机器学习如何使用远程标签

Question

I'm trying to use the firebase machine learning for the first time on flutter and Ive managed to upload the.tflite file in the custom model section easily enough and then use the remote model fine, the only issue I dont understand is if I update the model and want to add more labels I would ideally also use a remote.txt file but how would I upload that in the machine learning section as I can only find documentation showing how to use the model file itself.

FirebaseModelDownloader.instance
    .getModel(
        "Breed-Detector",
        FirebaseModelDownloadType.localModel,
        FirebaseModelDownloadConditions(
          iosAllowsCellularAccess: true,
          iosAllowsBackgroundDownloading: false,
          androidChargingRequired: false,
          androidWifiRequired: false,
          androidDeviceIdleRequired: false,
        ))
    .then((customModel) async {
  final localModelPath = customModel.file;

  // ...
  Tflite.close();
  String res;
  res = (await Tflite.loadModel(
      model: localModelPath.toString(), labels: "assets/labels.txt"))!;

so how would I use a remote labels.txt as well with the remote model rather than use a remote model and have to store the labels locally

Answer 1

I managed to use Firebase Storage for this, it miss versionning thought...

Here is my solution First create a bucket and update your file there: https://firebase.google.com/docs/storage/flutter/start

I used following rules:

rules_version = '2';
service firebase.storage {
  match /b/{bucket}/o {
    match /{allPaths=**} {
      allow read;
      allow write: if request.auth != null;
    }
  }
}

Then in flutter app:

late File labelsFile;

Future<File> getFirebaseFile(String fileName) async {
  Directory appDocDir = await getApplicationDocumentsDirectory();
  File labelsFile = File('${appDocDir.path}/$fileName');
  if (!labelsFile.existsSync()) {
    await downloadFirebaseFile(labelsFile, fileName);
  }
  return labelsFile;
}

Future<void> downloadFirebaseFile(File file, String fileName) async {
  try {
    await FirebaseStorage.instance.ref(fileName).writeToFile(file);
  } on FirebaseException catch (e) {
    print("Error loading labelsFile $e");
  }
}

Then you can use it right after localModelPath instanciation:

labelsFile = await getFirebaseFile("labels.txt");

// init inference
Tflite.loadModel(
    numThreads: numThreads,
    isAsset: false,                    //<-- this is important
    outputRawScores: outputRawScores,
    inputType: inputType,
    model: localModelPath.path,
    label: labelsFile.path,
);

And in the fetchLabels method:

Future<List<String>> fetchLabelList() async {
  List<String> labelList = [];
  labelsFile = await getFirebaseFile("labels.txt");
  await labelsFile.readAsString().then((q) {
    for (String i in const LineSplitter().convert(q)) {
      labelList.add(i);
    }
  });
  return labelList;
}

label's file will be downloaded if it does not exists localy, it just wont be updated if it changes version... Idk yet how to handle this...