Publication

Abstract : During ancient Indian times, the Gurukul system of education was the style of learning in the country. In this system the students were learning with their mentors (Gurus) and receiving education, knowledge, moral values and life skills under the guidance of their gurus. This system of education was practiced in ancient times, where all students who resided at the place of the guru in the Gurukula were considered equal. As the days passed the things took a drastic change and government schools were introduced. A few decades ago, the number of students in government schools decreased due to the privatization of the education system. Later, more schools were introduced by corporations, which led to a significant increase in the number of schools. To perform the tasks like monitoring and taking the attendance of each class was tedious job at the same time it was time consuming, where the total number of students in each class has increased and number of subjects for each class also increased and every teacher has to document the number of students attending the classes for each subject and they need to submit it to the higher authorities. To overcome these difficulties, In this paper a novel approach is schemed to record attendance of all the students of a class. In the proposed system to take the attendance of the students all at once, a live video is processed for each frame and to recognize the faces of all the students it uses a deep convolutional neural network face recognition algorithm which extract 128-dimensional face encodings from images and then compares these encodings with the faces stored in the dataset to determine the best match and further the attendance of the students present in the class is recorded in the form of excel sheet so that the teacher can carry out the further analysis. The findings of the experiment overcomes the difficulties faced in the existing systems and eyewitnesses the furturistic transition in marking attendance.