We've just published a post on using TensorFlow.Mobile from Android apps for on-device image classification using the Azure Custom Vision Service.
Check it out and feel free to discuss here!
I'd like to use the tensorflow object detection api to not only identify that a picture or video frame contains an object but to determine where in the frame the object exists. When do you expect that azure Custom vision service will have this capability?
Object detection is in preview at the moment. I don't know the timescales for when it will be generally available, but I imagine when that happens you will be able to export the models and use them on device.
Xamarin Inc., as a wholly-owned Microsoft subsidiary acting as a separate legal entity, adheres to the Microsoft Privacy Statement: Privacy & cookies