<![CDATA[The rapid digitalization era has driven the development of Artificial Intelligence (AI) technology, including machine learning approaches, as solutions for various challenges. In this research, the use of machine learning technology is aimed at providing beneficial innovations, especially in the world of higher education. This study aims to design and develop a face recognition-based attendance recognition system by integrating the YOLO (You Only Look Once) and K-NN (K-Nearest Neighbor) algorithms to optimize performance. Additionally, this research aims to implement real-time email attendance information delivery to students. Test results demonstrate that the YOLO algorithm can detect faces with high accuracy, reaching 99% - 100%, even in various situations such as changes in position and facial expressions, as well as the use of accessories like glasses Testing with variations in the K value and the number of training images in the K-NN algorithm yielded optimal results with a K value of 5 and 10 training images. The use of a K value of 5 achieved correlation and cosine distance calculations with an accuracy rate of 99% and highly efficient computational times ranging from 1.00 to 3.24 milliseconds (ms). Consequently, this system is proficient in recognizing faces with great accuracy and rapid responsiveness. Moreover, the system effectively delivers attendance information to students via email in real-time.]]>
Copyrights © 2025
