Research on Classroom Pose Recognition Algorithm Based on Deep Learning

In this paper, Self-Supervised Graph Media (SSGM), a class pose recognition algorithm combining MediaPipe, Graph Convolutional Network (GCN) and selfsupervised learning, is proposed. First of all, MediaPipe is used to efficiently extract key information of human body from video data to provide basic data for subsequent graph modeling. Secondly, GCN was used to model the graph structure of the extracted key data, capture the spatiotemporal dependence of human posture, and further optimize feature extraction. Finally, a self-supervised learning mechanism is introduced, and a large amount of unlabeled data is used for pre-training in an unsupervised way to enhance the robustness and generalization ability of the model to attitude changes. The experimental results show that SSGM algorithm performs well in many classroom pose recognition tasks. Compared with traditional methods, SSGM algorithm can capture students' pose information more accurately, optimize recognition strategies, and provide powerful technical support for teaching evaluation and behavior analysis. This achievement not only proves the effectiveness and feasibility of SSGM architecture in classroom posture recognition, but also provides a new idea for the intelligent development of educational technology.