This is an app demonstrating how to use CoreML and TensorFlow with Xamarin to build an image detection app. This app also uses the Xamarin.Forms Shell.
The model it uses was trained with Microsoft's Custom Vision service.
-
Create and deploy Azure Functions Project and add the url and keys to Constants.cs
-
All analysis is done using Custom Vision services. To use this solution, you must setup a model and then fill in the API Keys inside of Constants.cs
-
All the interesting code is in CustomVisionAPIService.cs. You can find the interfaces where we have wrapped the Custom Vision API.
-
For the Vision Models, the interface used by Xamarin.Forms can be found in ICustomVisionService.cs. Each platform level implementation can be found in their respective folders. For iOS, which uses CoreML can be found here and for Android, which uses TensorFlow can be found here