<![CDATA[Currently, there is a rising interest in utilizing the Internet of Things (IoT) for Smart home systems. One crucial aspect of Smart home systems is their security capabilities, specifically the ability to conveniently lock and unlock doors or gates. The primary issue in smart home security systems lies in their low accuracy and image processing delays, which were observed to be approximately 65% - 70% in experiments conducted using the KNN and Decision Tree methods. This research proposes a Deep Learning approach that achieves an accuracy of over 80%. The methodology employed in this study consists of four key steps: 1. Conducting a literature review on Smart Home Security, 2. Developing an RNN model for face detection, 3. Creating a prototype for face detection in a smart home setting, and 4. Evaluating the developed prototype for smart homes. The experimental results demonstrate that the proposed prototype achieves an accuracy of 94.3%. Furthermore, the recall rate is 94.3%, the f1 score is 91.66%, and the precision is 94.8%.]]>
Copyrights © 2023
