I have developed a custom AI model using Google AutoML Vision. I would like to build an app with AppInventor that uses the custom AI model. Is it possible ? If so, pls provide me with pointers to get started. I couldnt find relevant information on the internet.
I don't know what format(s) the AutoML Vision API provides, but if you can export the model into a format compatible with Tensorflow.js, we have built a number of example extensions using that API and the sources are available on GitHub under the Apache license. You could start with one of those extensions and wrap your model with it.
Thank You for the response. AutoML Vision provides Python client API and REST API to get predictions from the custom model. Appreciate if there is a way to use Python APIs.
There's no way in core App Inventor to use Python code. Android is developed in Java and so that's the lingua franca of most parts of the system. It may be possible with an extension to use Jython, a Java-based interpreter for Python, to do what you want. Alternatively, you could host a web-based API using a toolchain like Flask and talk to it with the Web component.