Scholarlink

<p>This paper presents EduVision Analyzer, a multimodal end-to-end system that bridges student engagement analytics with automated task management to address United Nations (UN) Sustainable Development Goal 4 (SDG 4 – Quality Education). The system accepts post-class video, audio, and live webcam input, applying computer vision for affect estimation, physical activity tracking, and eye-state monitoring, alongside Natural Language Processing (NLP) for lecture summarisation and actionable task extraction. A key architectural contribution is the direct integration with task management tools: action items inferred from lecture transcripts are automatically synchronised to student Trello boards, creating a closed-loop accountability mechanism. Unlike prior work focused solely on engagement monitoring, EduVision Analyzer connects analytics to structured, downstream student support. The affect estimation module employs a computationally lightweight heuristic proxy rather than deep emotion recognition, and all performance metrics are presented as proof-of-concept indicators from a 14-participant pilot; the limitations of this approach are discussed in full. The system makes a practical contribution toward SDG 4 indicators 4.1.1 (learner proficiency) and 4.4 (digital skills) by providing instructors with real-time engagement data and equipping students with structured task accountability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multimodal student engagement analytics with automated task management for SDG 4 aligned quality online education

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月20日

查看全文
  • K. Suthendran,
  • P. Linu Zenas Paul

摘要

This paper presents EduVision Analyzer, a multimodal end-to-end system that bridges student engagement analytics with automated task management to address United Nations (UN) Sustainable Development Goal 4 (SDG 4 – Quality Education). The system accepts post-class video, audio, and live webcam input, applying computer vision for affect estimation, physical activity tracking, and eye-state monitoring, alongside Natural Language Processing (NLP) for lecture summarisation and actionable task extraction. A key architectural contribution is the direct integration with task management tools: action items inferred from lecture transcripts are automatically synchronised to student Trello boards, creating a closed-loop accountability mechanism. Unlike prior work focused solely on engagement monitoring, EduVision Analyzer connects analytics to structured, downstream student support. The affect estimation module employs a computationally lightweight heuristic proxy rather than deep emotion recognition, and all performance metrics are presented as proof-of-concept indicators from a 14-participant pilot; the limitations of this approach are discussed in full. The system makes a practical contribution toward SDG 4 indicators 4.1.1 (learner proficiency) and 4.4 (digital skills) by providing instructors with real-time engagement data and equipping students with structured task accountability.