Student Motivation Analysis Based on Raising-Hand Videos

Area of Science:

• Educational Technology
• Human-Computer Interaction
• Artificial Intelligence in Education

Background:

• Current smart classroom research primarily focuses on recognizing hand-raising actions, neglecting detailed movement analysis.
• This oversight limits teachers' ability to leverage student behavior for enhanced pedagogical strategies.
• Assistive teaching methods require systems capable of both recognizing and analyzing hand-raising movements for targeted guidance.

Purpose of the Study:

• To propose an innovative morphology-based analysis method for detailed interpretation of student hand-raising movements.
• To address the limitations of existing deep learning methods in analyzing classroom hand-raising enthusiasm and creating behavioral databases.
• To enable smart classroom systems to provide targeted guidance based on a comprehensive understanding of student hand-raising behavior.

Main Methods:

• Utilized a neural network, specifically YOLOX (object detection) and HrNet (skeleton estimation), to obtain student skeleton key point data.
• Developed a morphology-based analysis to convert skeleton key point data into several one-dimensional time series.
• Analyzed these time series to extract detailed information on the speed and amplitude of hand-raising movements.

Main Results:

• Successfully recognized hand-raising actions with a high degree of accuracy.
• Provided a detailed analysis of hand-raising movement speed and amplitude, surpassing the capabilities of coarse neural network recognition.
• Demonstrated the effectiveness of the proposed method through experimental validation.

Conclusions:

• The morphology-based analysis method offers a significant advancement in understanding student behavior within smart classrooms.
• This approach effectively supplements existing deep learning models by providing richer, actionable insights into student engagement.
• The findings pave the way for more responsive and personalized assistive teaching strategies in intelligent educational environments.