How to detect object on the image with oxxnruntime_flutter using a pre-trained *.onnx model?

[frg-x]

Please give an example how to inference image with pre-trained *.onnx model using your package?
At the beginning I've trained YoloV8 model and got *.pt trained model, then exported it to *.onnx format with opset13.

Inputs: [name: 'images', tensor: float32[1, 3, 320, 320]]
Outputs: [name: 'output0', tensor: float32[1, 6, 2100]] 

[1, 6, 2100] in Outputs - 6 means TWO objects to detect, 5 means ONE objects to detect

I've tried inference in Python with onnxruntime package and it works great, but I can't get any results with your package. How to read/prepare image bytes to pass them as argument to createTensorWithDataList method?

Here is my chunk of a code:

Int16List _transformBuffer(Uint8List bytes) => Int16List.view(bytes.buffer);

Future<void> inference() async {

    final List<int> shape = [1, 3, 320, 320];

    final floatBuffer = _transformBuffer(imageBytesList).map((e) => e / 255).toList();

    array = Float32List.fromList(floatBuffer);

    final OrtValueTensor inputOrt = OrtValueTensor.createTensorWithDataList(array, shape);

    final Map<String, OrtValueTensor> inputs = {'images': inputOrt};

    final runOptions = OrtRunOptions();

    final List<OrtValue?>? outputs = await session.runAsync(runOptions, inputs, ['output0']);
}

Thank you in advance