Selma is virtual assistant can help you to control all kind of appliances by using bluetooth low energy and also you have to know that this virtual assistant is backed by augmented reality capabilities to allow the user to interact natively with his home.
in this project we have natural language processing within an android device by building three blocks :
speech to text capability
natural language understanding using Dialogflow
test to speech capability
also we have built an Augmented reality capabilities by using real time object recognition we have chosen SSD model which is implemented by tensorflow the model generated from Tensorflow has been converted by using tflite to transofom it to model able to be run on android device .
to control our appliance the smart home should be connected to raspberry pi 3 and beacon nr51822 to control lamp TV and air conditioner over bluetooth low energy