尊敬的
微信汇率：1円 ≈ 0.046166 元
支付宝汇率：1円 ≈ 0.046257元
[退出登录]
雅虎竞拍
煤炉代买
paypay商城
乐天二手
日本亚马逊
乐天新品
ZOZOTOWN
This document discusses a proposed system to monitor student engagement in online learning environments using face detection. The system would use face recognition and head pose estimation to authenticate that students are present and attentive during online lectures. Student engagement is important for learning outcomes, but more difficult to monitor online compared to in-person. The proposed system would collect data on attention, emotions, and activities to provide insights on class and student engagement levels. This could help instructors evaluate their teaching methods and identify students who may need extra support. The document outlines the implementation of this system using tools like the DAiSEE dataset for emotion detection and analyzing head pose to estimate attentiveness. It also provides examples of what the instructor dashboard and student interfaces may look like.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 10 Issue: 12 | Dec 2023 www.irjet.net p-ISSN: 2395-0072
© 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 131
progress gives teachers the opportunity to reflect on
their teaching and to evaluate the effect of the teaching
methods they are using. Another major benefit of
monitoring student progress is that it allows the
teacher to evaluate the effectiveness of his own
teaching. If the rest of the class finds it difficult to
understand or express a particular purpose, it may not
be the students' ability to solve the problem.Asa result,
teachers can help students reach their educational
goals.
With information from assessment and activity
samples, the teacher can work with the student to
establish learning goals that are achievable and help
each student stay on track. With continuous student
supervision, teachers can establish an achievable and
individual level of progress for each learner, or vice
versa and intervene where necessary.
2. RELATED WORK
[1] A method for detecting learner participation has been
created using the Local Directional Pattern (LDP) and the
Deep Belief Network (DBN). In each frame, facial
characteristics were extracted as LDP-histograms by
applying a fixed mask to the identifiedfaceregionofinterest.
During classification, involvement detection judgements
were conducted at twolevels(notengagedandengaged)and
three levels (not engaged, usually engaged, and extremely
involved).
[2] For engagement detection, we leverage behavioral and
emotional data from students. All Convolutional Network,
Network in Network, Very Deep Convolutional Network,
Conv-Pool Convolutional Network, and a proposed model
combining some special features from the above models
have been tested for behavioral and emotional dimensions
detection to detect student engagement in online learning.
[3] During a lesson, a web application was created to
facilitate various sorts of student-teacher interaction. The
programme collected all of the auditory exchanges and
labelled the data so that further analysis of the data could be
done quickly.
[4] The open-source datasetwasusedtodemonstratea web-
camera-based system for evaluating the possibilities of
automated student engagementmonitoringandassessment.
A case study was conducted to get insight into the genuine
challenges of student engagement detection and
identification when students are allowed to sit and move
freely during learning. The student images in the datasetare
used to train a deep learning CNN model, which is then used
for real-time monitoring and assessment of student
involvement.
[5] The idea is to make existing facial recognition systems
more accurate. From still pictures and video frames, face
recognition is possible.
[6] It monitors keyboard, mouse, and clickstream data. A
webcam for estimating head pose. Browser tab activation,
upgrades and deletions, and website screenshots are all
possible. Listener for browser information and window
focus.
[7] It measures the engagement of the studentthroughfacial
recognition by detecting various features of students face
such as head pose, facial fiducial points,learnedfeatures,eye
gaze, etc.
[8] It depends on head stance and basic facial gestures like
brow raise, eye closure, and upper lip rise to form their
judgments about student interest.
[9] To achieve its purpose, it applied machine learning
approaches to educational data acquiredina hybridlearning
environment. The studyfocusedonthecorrelationsbetween
engagement level andstudentacademicachievement,aswell
as how machine learning algorithms could assist educators
in automatically monitoring and responding to students'
learning progress concerns, allowing them to focus on other
pedagogical issues.
3. PROPOSED SYSTEM
This System will identify whether the student is attending
the lecture or not, and the tool automatically registers the
time he/she is in class (class start time relative to the
system's clock). It also gathers his or her head position. Our
aim will be to have a score of attentiveness and an average
score per each student. Based on the results we will begiven
feedback, recommend activities and observe the student's
behavior. We will also be monitoring mouse and keyboard
clicks. If the student is not attending, we will notify the
instructor. We will also be tracking some of the elements
such as browser tabs, website snapshots and other activity.
In the end we will be making a dashboard for instructors to
present to their students with their average student
engagement levels. The dashboard will contain a lot of
valuable information for instructorssuchasclasseswithlow
engagement levels, students who are not attending lectures
etc. The information collected from our tracking system can
be used in many different ways suchasimprovingclassroom
or lecture training and how to improvestudentperformance
i.e., what questions should be asked during each lecture,
what questions should be asked after each lecture etc.Andit
will authenticate students via facial recognition and head
pose estimation.](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
