Development of an Automatic Class Attendance System using CNN-based Face Recognition

Systematic Literature Review: Biometric Technology Applied to Educational Institutions

“…The system can mark multiple attendances from a single group image captured by a video surveillance camera. [31] In the research, face recognition is performed using convolutional neural networks with data input, dataset training, face recognition, and attendance input to develop an automatic attendance system. The application detected and recognized students in the class with a maximum accuracy of 92%.…”

“…Related studies Prevent impersonation and false assistance [21], [22], [27], [28], [30], [31], [32], [34], [35], [38], [39], [40], [41] Saving time and work for teachers and students [21], [23], [33], [34], [37], [38], [41], [43] Increasing the security and reliability of data [21], [24], [25], [26], [27], [28], [29], [35], [36], [37], [38], [42] Increases productivity [23] Improving work culture, accountability, and transparency [23], [40], [41], [43] Generates a clear audit trail for attendance records [23] Biometric traits do not change or are difficult to alter over time.…”

Emotion Detection Using Machine Learning: An Analytical Review

“…on the other hand, build a bullying detection algorithm based on speech emotion recognition and motion recognition to accurately identify bullying incidents in real-time is the major goal of an article [16]. A study paper suggests an automated attendance system that makes use of face recognition technology, which is an innovative concept in the field [17]. An article applying deep learning techniques to detect emotions from facial expressions also uses an AI system, with the overall goal of improving the effectiveness of current models and making a contribution to the field [18].…”

“…Each user trained a specific number of images, and the precision was evaluated by using the system to separate the trained images from five different video source frames. In the training phase, they experimented four times while increasing the number of photos per person [17]. Next, only the video-based emotion recognition subchallenge was completed by Fan et al Each film in this collection is categorized into one of seven emotions: disgust, fear, happiness, sadness, surprise, and neutral.…”

“…The camera used for face recognition should be positioned in a way that it has a clear view of all the students to ensure accurate detection and recognition. The system may label a student as 'Unknown' if there is not a match of high enough accuracy during face recognition [17]. The constraints of the deep learning technique are not explicitly discussed in the research on negative emotion identification in Thai.…”