Title: Deep learning for smart home security systems: CNN-based home security
Authors: M. Balasubramanian; Kiran Shrimant Kakade; Sulakshana B. Mane; Sujatha Jamuna Anand
Addresses: Department of Computer Science and Engineering, S.A. Engineering College, Chennai, Tamil Nadu, India ' Faculty of Business and Leadership, MIT, World Peace University, Pune, India ' Bharati Vidyapeeth College of Engineering, Navi Mumbai, India ' Department of Electronics and Communication Engineering, Loyola Institute of Technology, Chennai, India
Abstract: A smart home enables new modes of connection and the consumption of various services. Additionally, AI and deep learning have aided in the enhancement of many services and jobs by making them more automated. In this study, we used IoT and deep learning to create a safe and efficient home automation system. Using deep learning approach, the user is able to operate appliances such as fans, televisions, bulbs, and other electronic or electrical equipment by either speaking commands into their mobile device or using an application that is pre-installed on their mobile device. The results of the trials that were carried out demonstrate that the suggested deep learning model is more accurate than the KNN method, and that the RL system improves the user's quality of experience by as much as 3.8 points on a scale of 10.
Keywords: deep learning; smart home; IoT; RL.
DOI: 10.1504/IJESDF.2024.139652
International Journal of Electronic Security and Digital Forensics, 2024 Vol.16 No.4, pp.451 - 463
Received: 26 Oct 2022
Accepted: 14 Dec 2022
Published online: 05 Jul 2024 *