International Research Journal of Engineering and Technology (IRJET) Volume: 09 Issue: 11 | Nov 2022
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
A Review on Students Attention Monitoring Systems Parvathy G S 1, Smitha R 2 1Lecturer, Dept. of
Computer Engineering, NSS Polytechnic College, Pandalam, Kerala Electronics and Communication Engineering, NSS Polytechnic College, Pandalam, Kerala ---------------------------------------------------------------------***--------------------------------------------------------------------been presented in [2]. Then, Continuous performance tests Abstract –Blackboard lectures with slides are still a popular 2Lecturer, Dept. of
(CPTs) are used to collect attention responses from students and their corresponding EEG signals. Using NeuroSky's brainwave detector and the support vector machine (SVM), the AAS is constructed. The recorded EEG signals are then divided into five major bands (alpha, beta, gamma, theta, and delta), each of which has five statistical parameters, using discrete wavelet transforms (DWT). Based on twenty-five unique brainwave properties related to attention levels, the AAS is constructed. The proposed Attention awareness system has a 90.39% accuracy rate.
teaching method. The fact that lectures frequently keep learners in a passive state is one issue with this method. Passive listeners may find it challenging to stay focused and pay attention throughout a lecture. This can make the lecture less effective in terms of learning results. This approach does not provide teachers with real-time feedback. The ability to assess the attention of students is the foundation for improving both instructional delivery and student learning. A teacher has difficulty tracking and monitoring every student's interest in learning, especially in a large classroom. This study reviews the various currently existing student attention monitoring systems, that automatically monitors students’ attention level and gives live feedback to teacher.
An attention-tracking system that just requires a simple web camera in real-time is developed in [3]. In addition to scale and rotation invariant, this system is also tolerant to incorrect attention state classifications brought on by blinks. The facial region is identified by using Haar Cascade Classifier. Using a Haar Cascade Classifier tailored for eye identification, the positions of both eyes are identified. Following the detection of the inner and outer corner points of the eyes in a frame, the Lucas-Kanade tracker will detect the corner nodes in consecutive frames. The average intensity in the ROI region following the subtraction procedures is used to assess attention. A feedback channel is incorporated in an active attention-tracking system to warn the user with alarm signals when inattentiveness is identified. The database will keep the attention status and the accompanying bookmarks for the E-learning content.
Key Words: Face Detectors, Discrete Wavelet Transform (DWT), Support Vector Machine (SVM), Haar Cascade Classifier, Convolutional Neural Network (CNN), Face Alignment Network, Electroencephalograph(EEG), Kinect Sensor, Histogram of Gradient (HOG) descriptor, Fisherface Algorithm
1. INTRODUCTION In traditional face-to-face instruction, teachers generally observe students’ facial expressions to determine whether they are sufficiently attentive. However, this method is excessively subjective and consumes a significant amount of the teacher’s energy. In addition to face-to-face instruction, e-learning allows However, students may become easily distracted in e-learning environments, owing to the absence of a teacher’s face-to-face supervision Despite the importance of maintaining sustained attention during a learning activity to ensure successful learning, evaluating whether students maintain their concentration on a learning activity is extremely difficult, owing to the lack of supervised mechanisms to monitor their attention states [1].
The method of assessing student attention spans through facial landmarks was proposed in [4]. Lips and eyes on the face are highlighted here as areas of special importance.. Aspect ratios for the eyes and mouth were calculated, as well as threshold values. A face detector that combines the Histogram of Oriented Gradients (HOG) and Linear Support Vector Machine (SVM) techniques were used find the student's face. Student states were determined based on the coordinates of the eyes, mouth, and face landmarks: 'Normal', 'Dozing', or 'Yawning'. The basis for yawning, nodding off, and paying attention is a fixed number of consecutive frames that shouldn't be exceeds the maximum. A display showing the determined states is directly visible to students and instructors.
This article examines different student attention level assessing systems. In Chapter 2, the various methods of assessments are discussed. The comparison of the various approaches is shown in Chapter 3. The study's conclusion is presented in Chapter 4.
In [5], students' level of concentration was assessed using a deep learning approach. The system utilizes two types of facial cues to determine student concentration levels. They include gaze and fatigue detection. Consistent identification
2. LITERATURE REVIEW Using raw human EEG data captured by Neurosky's Mind Wave earphone, a novel attention-aware system (AAS) has
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 688