Scholarlink

<p>To enhance the effectiveness of AI-enabled smart classroom teaching behavior monitoring, a method based on the internet of things (IoT) is proposed. This paper discusses about the design of a monitoring framework that utilizes IoT radio frequency identification (RFID) technology to capture student behavior signals, recover the true phase trajectory through phase unwrapping processing, and then filter and smooth the phase and received signal strength indicator (RSSI) signals to reduce noise and obtain teaching behavior data and perform category imbalance processing. Long short-term memory (LSTM) and convolutional long short-term memory (ConvLSTM) models were used to extract video temporal and spatial features, respectively. Based on transfer learning techniques, complete feature extraction was done and a classification loss function was constructed to solve the monitoring results. The results showed that this method achieved an accuracy of up to 99% in monitoring teaching behavior abnormalities, with an abnormality capture delay of no more than 40 ms. This method can accurately monitor teaching behavior and promptly detect abnormalities.</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

AI-empowered smart classroom teaching behavior monitoring method based on the internet of things

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • Fei Liu

摘要

To enhance the effectiveness of AI-enabled smart classroom teaching behavior monitoring, a method based on the internet of things (IoT) is proposed. This paper discusses about the design of a monitoring framework that utilizes IoT radio frequency identification (RFID) technology to capture student behavior signals, recover the true phase trajectory through phase unwrapping processing, and then filter and smooth the phase and received signal strength indicator (RSSI) signals to reduce noise and obtain teaching behavior data and perform category imbalance processing. Long short-term memory (LSTM) and convolutional long short-term memory (ConvLSTM) models were used to extract video temporal and spatial features, respectively. Based on transfer learning techniques, complete feature extraction was done and a classification loss function was constructed to solve the monitoring results. The results showed that this method achieved an accuracy of up to 99% in monitoring teaching behavior abnormalities, with an abnormality capture delay of no more than 40 ms. This method can accurately monitor teaching behavior and promptly detect abnormalities.