Tflite custom object detection model for flutter application using google mlkit object detection package

Question

I'm trying to create a custom object detection model in tflite format so that I can use it in a flutter application with the google_mlkit_object_detection package.

The first few models, I created using yolov8 and converted to tflite. The annotations were made with roboflow and the google colab notebook I used was provided by them, the metadata of the converted tflite model looks like the following image

On this model I was getting the error

Input tensor has type kTfLiteFloat32: it requires specifying NormalizationOptions metadata to preprocess input images.

So as suggested I tried to change the metadata and add normalizationOptions but failed to do so. My second alternative was to train a model with the official TensorFlow google colab notebook TensorFlow Lite Model Maker and it generated a model with the following metadata

For this model the error was

Unexpected number of dimensions for output index 1: got 3D, expected either 2D (BxN with B=1) or 4D

So I checked the model from the example app from the package I am using "google_mlkit_object_detection" and the metadata looks like this

So my question is, how can I alter the models I already trained whichever it is easier, to look like this, both input and output, do I have to alter my model's architecture or just the metadata? The second one trained with the official notebook from tensor flow, it seems that all I have to do is include the correct shape format [1,N], but again I might have to change the architecture.

Answer 1

It seems like the issue is with the input and output shapes of your custom object detection model, which do not match the expected shapes of the google_mlkit_object_detection package.

You can try modifying your model's architecture to match the expected input and output shapes. Alternatively, you can modify the metadata of your existing models to match the expected shapes, but this may require some trial and error.

To modify the metadata, you can use the TensorFlow Lite Metadata Writer API or a metadata editor tool such as Netron. You would need to update the metadata to include the correct input and output shapes, as well as any other required information such as normalization options.

Once you have updated the metadata, you can use the google_mlkit_object_detection package to load the model and perform object detection in your Flutter app.

To know more about Flutter, join our Flutter APP Development Course today.