Is there a way to implement batch classification in Flutter with Tflite?

Question

I want to classify multiple images iteratively, I tried it with

  Future pickImages() async {
final List<XFile>? images = await ImagePicker().pickMultiImage();
if (images == null) return;


images.forEach((image) async {
  print('- Name: ${image.path}');
  await runModel(image.path);
});

}

Future runModel(String imagePath) async {
var recognitions = await Tflite.runModelOnImage(
    path: imagePath,
    numResults: 1,
    threshold: 0.5,
    imageMean: 127.5,
    imageStd: 127.5,
    asynch: true // defaults to true
    );

print(recognitions);
await Future.delayed(Duration(seconds: 1));
setState(() {});

}

Everything runs fine if I just select one image, as soon as I go on selecting multiple images I get the following error:

PlatformException (PlatformException(Failed to run model, Interpreter busy, java.lang.RuntimeException: Interpreter busy

I tried my best but couldn't resolve the issue.

I would really appreciate it if you can suggest how to make inferences on multiple selected images(ImagePicker).

Thanks!

Answer 1

this worked for me:

for pickImage:

async {
final ImagePicker _picker = ImagePicker();
final pickedFile = await _picker.pickMultiImage();
List<XFile> xfilePick = pickedFile;
setState(
      () {
    if (xfilePick.isNotEmpty) {
      for (var i = 0; i < xfilePick.length; i++) {
        selectedImages.add(File(xfilePick[i].path));
      }
    }
  },
);


for (var i = 0; i < selectedImages.length; i++){
    await Future.delayed(const Duration(seconds: 1));{
      imageClassification(selectedImages[i]);
      await Future.delayed(const Duration(seconds: 1));
    }
  }}