On-Device Mobile Inferencing Using CoreML, Tensorflow Android Inference, and Xamarin.Forms
Companion code for the article Delivering On-Device AI Solutions Using Xamarin, CoreML, and Tensorflow Android Inference
This is an example program showing the delivery of a Tensorflow-trained LSTM on both iOS, using CoreML, and Android, using Tensorflow Android Inference.
This Visual Studio for Mac solution contains all necessary files, including the trained ML model and Xamarin binding project for Tensorflow Android Inference. The binding project used by the sample is located in this repo under TensorFlowXamarin/ but developers interested in the latest version of the binding should refer to the Tensorflow.Xamarin.Android dedicated repository.
The /training directory contains data and Python sources for recreating the model and converting/saving the Keras code into CoreML and Tensorflow protobuf format.
Just in case it needs to be made explicit: Under no circumstances should you use the data or model here for planning or navigation!