Applied Mathematics and Nonlinear Sciences, Journal Year: 2024, Volume and Issue: 9(1)
Published: Jan. 1, 2024
Abstract This paper proposes a method for recognizing teacher behavior patterns using graph convolutional neural network. extracts the human skeleton point information by single-frame image from teaching video as unit. Then, network is used framework to aggregate multi-frame and identify teacher’s behavioral patterns. In order verify effectiveness of method, article constructs dataset collected public music videos on Catechism.com, which contains six categories conducts comparative experiments. The experimental results show that pattern recognition technique based can effectively exclude interference irrelevant within classroom scene make full use generated between points in accurately typical behaviors teacher, with an accuracy more than 0.96 stronger robustness. study respond teachers’ status timely effective manner, help teachers optimize their improve classrooms.
Language: Английский