Scholarlink

<p>This paper presents a novel cybernetic framework for real-time student behavior classification in classroom environments. The system integrates machine learning with cybernetic principles to create an adaptive monitoring system capable of intelligent data filtering and multimodal feature processing. A key innovation is the intelligent filtering mechanism positioned between detection and tracking stages, which functions as a cybernetic feedback controller to minimize computational overhead while maintaining accuracy. The system processes facial features via Py-Feat and body pose features through OpenPose, creating a comprehensive behavioral representation. Experimental results demonstrate 93.03% classification accuracy with 70.03% reduction in detection time and 20.30% decrease in classification time compared to baseline methods. The cybernetic framework enables real-time deployment in educational environments while providing adaptive system behavior based on environmental feedback to improving teaching performance, supporting learning processes, enhancing student achievement, and learning outcomes.</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

Cybernetics-based filtering for real-time student behavior classification

  • 期刊论文

  • 发表日期：

    2026年7月22日

查看全文
  • Xuan Hieu Vuong,
  • Duc Trung Mai,
  • Hong Hai Trinh,
  • Trung Kien Phuong,
  • Nhat Hai Nguyen,
  • Tuan Linh Dang

摘要

This paper presents a novel cybernetic framework for real-time student behavior classification in classroom environments. The system integrates machine learning with cybernetic principles to create an adaptive monitoring system capable of intelligent data filtering and multimodal feature processing. A key innovation is the intelligent filtering mechanism positioned between detection and tracking stages, which functions as a cybernetic feedback controller to minimize computational overhead while maintaining accuracy. The system processes facial features via Py-Feat and body pose features through OpenPose, creating a comprehensive behavioral representation. Experimental results demonstrate 93.03% classification accuracy with 70.03% reduction in detection time and 20.30% decrease in classification time compared to baseline methods. The cybernetic framework enables real-time deployment in educational environments while providing adaptive system behavior based on environmental feedback to improving teaching performance, supporting learning processes, enhancing student achievement, and learning outcomes.