Asis, Jaymark Laurence D.; Barceleno, Jacob E.; Mariano, Marc Ernest C.
IoT-Powered Baby Care: enhancing safety and convenience with a smart baby cradle featuring face orientation detection using multi-task cascaded convolutional neural networks (MTCNN) - Undergraduate Thesis: (Bachelor of Science in Information Technology) - Pamantasan ng Lungsod ng Maynila, 2024.
ABSTRACT: This study addresses the growing need for advanced infant care solutions by proposing the design and development of an Internet of Things (IoT)-enabled smart baby cradle. Focused on enhancing infant safety and well-being, especially in relation to reducing the risk of Sudden Infant Death Syndrome (SIDS), the system incorporates Multi-task Cascaded Convulutional Neural Network (MTCNN) technology for neonate face detection and facial feature alignment. The smart cradle integrates environmental sensors, a speaker system for soothing sounds, and a motorized system for comforting movements, all controlled by a Raspberry Pi 4 Model B. Following the Waterfall Development Methodology, the research undergoes Requirements Gathering, System Design, Implementation, Integration and Testing, Deployment, and Maintenance phases. Notably, this study uses assessed the device and application using ISO 25010:2011 standards, specifically focusing on functional suitability, usability, reliability, and portability. The results indicate high satisfaction with IoT-powered smart baby cradle, with mean scores above 3.26 (out of 4) for functional suitability, usability, reliability, and portability. Notably, the cradle received favorable ratings for swaying motion (3.514), features (3.629), safety and well-being (3.429), alert systems (3.514), automatic response (3.457), and portability features (3.314). The MTCNN model demonstrated robust face detection performance across varying levels of image complexity, achieving commendable area under the curve (AUC) values of 0.801, 0.771, and 0.526 for Easy, Medium, and Hard sets, respectively. Additionally, on the FDDB benchmark, the MTCNN exhibited a True Positive rate of 0.93 and an Average Precision of 0.901, showcasing its effectiveness in accurately identifying faces. The integration of infant face detection, tracking, and a Sleep Analysis module in this research showcases the system’s effectiveness in enhancing infant safety and sleep quality while providing parents with remote monitoring capabilities. This interdisciplinary approach underscores the potential of IoT technology to address safety concerns, improve caregiving experiences, and contribute to the advancement of infant care practices, emphasizing the importance of user-centered design for tailored solutions.
5
academic writing
T58.6 A85 2024
IoT-Powered Baby Care: enhancing safety and convenience with a smart baby cradle featuring face orientation detection using multi-task cascaded convolutional neural networks (MTCNN) - Undergraduate Thesis: (Bachelor of Science in Information Technology) - Pamantasan ng Lungsod ng Maynila, 2024.
ABSTRACT: This study addresses the growing need for advanced infant care solutions by proposing the design and development of an Internet of Things (IoT)-enabled smart baby cradle. Focused on enhancing infant safety and well-being, especially in relation to reducing the risk of Sudden Infant Death Syndrome (SIDS), the system incorporates Multi-task Cascaded Convulutional Neural Network (MTCNN) technology for neonate face detection and facial feature alignment. The smart cradle integrates environmental sensors, a speaker system for soothing sounds, and a motorized system for comforting movements, all controlled by a Raspberry Pi 4 Model B. Following the Waterfall Development Methodology, the research undergoes Requirements Gathering, System Design, Implementation, Integration and Testing, Deployment, and Maintenance phases. Notably, this study uses assessed the device and application using ISO 25010:2011 standards, specifically focusing on functional suitability, usability, reliability, and portability. The results indicate high satisfaction with IoT-powered smart baby cradle, with mean scores above 3.26 (out of 4) for functional suitability, usability, reliability, and portability. Notably, the cradle received favorable ratings for swaying motion (3.514), features (3.629), safety and well-being (3.429), alert systems (3.514), automatic response (3.457), and portability features (3.314). The MTCNN model demonstrated robust face detection performance across varying levels of image complexity, achieving commendable area under the curve (AUC) values of 0.801, 0.771, and 0.526 for Easy, Medium, and Hard sets, respectively. Additionally, on the FDDB benchmark, the MTCNN exhibited a True Positive rate of 0.93 and an Average Precision of 0.901, showcasing its effectiveness in accurately identifying faces. The integration of infant face detection, tracking, and a Sleep Analysis module in this research showcases the system’s effectiveness in enhancing infant safety and sleep quality while providing parents with remote monitoring capabilities. This interdisciplinary approach underscores the potential of IoT technology to address safety concerns, improve caregiving experiences, and contribute to the advancement of infant care practices, emphasizing the importance of user-centered design for tailored solutions.
5
academic writing
T58.6 A85 2024