Mobile Based Attendance System: Face Recognition and Location Detection using Machine Learning

Nowadays, student attendance has a significant impact on the class process as well as it affects student marks based on the authenticity of marking the attendance which expects high accuracy with more flexibility for both lecturers and students. Currently, most universities have some methods of marking student attendance, which consider either inefficient in terms of accuracy and security or bring up other issues such as time-consuming. In addition, another way of attendance requires an administrator to monitor the input of student data, which is difficult when dealing with a large number of students. This project proposes a mobile application to allow lecturers to generate class attendance and students to submit the attendance by scanning their faces using their mobile phone camera along with their location. The proposed system addressed three main components: First, automatic face detection and analysis using mobile phone cameras. Second, face recognition API that uses machine learning algorithm. Third, maps API. A testing process has been conducted through over 30 students to test the efficiency of the system in recognizing student faces as well as location. The result shows face recognition has achieved a high accuracy of detecting students’ faces even in a bad environment condition. It revealed that over 85% of the student are satisfied with the face recognition process. The system showed effective example responses by marking the student's attendance after recognizing the student's face and location, and the lecturer was able to download a report of submitted attendance.