Video-Based Student Engagement Estimation via Time Convolution Neural Networks for Remote Learning
- Publisher:
- SPRINGER INTERNATIONAL PUBLISHING AG
- Publication Type:
- Conference Proceeding
- Citation:
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, 13151 LNAI, pp. 658-667
- Issue Date:
- 2022-01-01
Closed Access
| Filename | Description | Size | |||
|---|---|---|---|---|---|
| 978-3-030-97546-3_53.pdf | Published version | 417.09 kB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
Given the recent outbreak of COVID-19 pandemic globally, most of the schools and universities have adapted many of the learning materials and lectures to be delivered online. As a result, the necessity to have some quantifiable measures of how the students are perceiving and interacting with this ‘new normal’ way of education became inevitable. In this work, we are focusing on the engagement metric which was shown in the literature to be a strong indicator of how students are dealing with the information and the knowledge being presented to them. In this regard, we have proposed a novel data-driven approach based on a special variant of convolutional neural networks that can predict the students’ engagement levels from a video feed of students’ faces. Our proposed framework has achieved a promising mean-squared error (MSE) score of only 0.07 when evaluated on a real dataset of students taking an online course. Moreover, the proposed framework has achieved superior results when compared with two baseline models that are commonly utilised in the literature for tackling this problem.
Please use this identifier to cite or link to this item: