Mobile Deep Learning with TensorFlow Lite, ML Kit and Flutter

Using an on-device TensorFlow Lite model

Now, let's add functionality for the second choice of the user—that is, analyzing the chosen image using a TensorFlow Lite model. Here, we will be using our previously created TensorFlow Lite model. The following steps discuss in detail how to use an on-device TensorFlow Lite model:

1. We will start by adding the tflite dependency in the pubspec.yaml file, like this:

dev_dependencies:flutter_test:  sdk: flutterimage_picker: ^0.6.0//Adding tflite dependencytflite: ^0.0.5

2. Next, we configure aaptOptions in Android. Add the following lines of code to the android/app/build.gradle file, inside the android block:

aaptOptions {       noCompress 'tflite'       noCompress 'lite'   }

The preceding code snippet makes sure ...