Configuration-Free Systems for WiFi Sensing based Smart Home using the Smart Remote Controller

Abstract

The Smart Home invention offers us extreme supervision over our home by automating the lighting structure, the dimming, the screens, electric machines, the sound and safety frames. Main technologies that provide connectivity to smart home facilities, WiFi is one of them. In traditional households, where the household appliances will increase, the remote controls to manage them and the interference between them will also increase. This makes the system configuration dependent and troublesome for users to manage them that increases their burden. Recently, some systems were developed to manage multiple household appliances through one interface. However, the matter is once the dataset will increases the interface gets sophisticated and every home appliance or controller wants a special device to connect with it. In this paper, we introduced FreeGesture in the DeepRemote controller, which is a gesture recognition scheme without a device that uses preferred computer vision algorithms, particularly deep learning, to recognize numerous devices and manage them via IR or network. It simplifies deployment and makes systems without configuration. We are going to consider it as “Smart Remote” in this article. Smart Remote consists of four buttons, a camera, an Inertial Measurement Unit (IMU), a WiFi component, an Infrared (IR) transceiver, and a speaker. The popularity accuracy of smart Remote for 5 varieties of home appliances from completely different places is hyperbolic from 81.07% to 95.8%.