DEEP LEARNING IMPLEMENTATION FOR SMART HOME SYSTEM BASED ON ARDUINO UNO

Abstract

Smart Home is a combination of technology and services in the home environment with the aim of increasing efficiency, comfort and safety. Most of the smart home systems nowadays, still need human intervention in its implementation. This thesis will develop a prototype of a Smart Home by using Deep Learning method, which has the ability in recognizing and identifying the face and expressions of the authorized person. There will be two main features in the application, including a security system with face recognition and controlling home appliances by using facial expression. This smart home system will use the Deep Convolutional Neural Network algorithm on TensorFlow with the Inception V3 for its Architecture model.