| dc.description.abstract | With the substantial growth of artificial intelligence and a big concern about property crimes
these days, topics about home security are also well-concerned. That is why we want to research
Face Recognition for University based on Smart Door Lock System. Home security can be used
for surveillance at home, business, and university, which will be focused on my thesis.
In this thesis, we aim to develop a face-recognition-based smart door lock system to improve
safety and security in school environments, save more time and effort for lecturers, security
guards, and also laborers not only to control the classroom but also to lock or unlock the
classroom’s door every single day before the beginning and after the end of the class in an
easier, faster and also more convenient way. In this report, we chose Multi-task Cascaded
Convolutional Networks (MTCNN) and Facenet Algorithm to be mainly proposed. Besides
our proposed algorithm, we also designed the demonstration for the hardware, which is Face
Recognition-based real-time smart door lock system using Raspberry Pi 4, and a Web Server,
which presents the results from the Hardware via MQTT communication, to ensure that the
door is only open for the right Lecturers and the Administrators such as MTCNN and Facenet. | en_US |
