Offline voice detection in smart homes

Smart cities, homes, automobiles, and the internet of things, in general,have changed human’s life. In this paper, we emphasize on smart homeswhich include automation, authentication, voice analysis, and interactingwith the environment. Also, a multi-purpose smart home automationsystem (SHAS) is proposed in which the users can make use of voicecommands to control connected devices and for authentication. Theproposed system can detect speech regardless of the accent, dialect, andemotions. The aim of implementing this system is to reduce theconsumed electrical energy, human resources, and equipment costs. Dueto the considerable energy consumption in homes, domestic energymanagement [۱] is considered an important and fundamental issue.Reducing electricity cost is carried out without considering dramaticalterations in the consumption pattern, and the domestic energyconsumption program is performed dynamically to reduce energyconsumption using environmental information. Voice commandsdetection and performing the commands are carried out by a hardwaremodule called Raspberry pie. As it is known, smart homes can be used inboth online and offline modes, but offline performance is of theadvantages of this paper, it means, offline performing of the voicecommands in a smart home is the aim of this paper. This advantage showsits superiority for protecting personal data privacy where the network isunavailable or the coverage is poor, which is common in regions far fromantennas. Also, due to the threat of communication between the processing unit, sensors and other components of the smart home, fromcyber attacks use a firewall to secure and achieve smart home goals.

Keywords:

Smart home , voice detection , home automation , Raspberry pie , cyberattack