Research And Application Of Human Motion Recognition On Immersive Chinese Phylosophy Education

Current convolutional neural network technology is driving the rapid development of computer vision technology,and human action recognition algorithms based on a series of video image technologies have become a popular topic of current research.Whether in security monitoring,elderly monitoring or in medical diagnosis and abnormal behaviour monitoring,human action recognition has potential research value and broad application prospects.With the high importance attached to family education in China,the use of intelligent technology to promote cultural education methods in the family environment has emerged,but most of the methods lack the ability to generate corresponding perceptual communication with cultural content.This paper therefore focuses on a human action recognition algorithm that incorporates attention mechanisms and applies it to a home scene with the intention of creating an immersive cultural education atmosphere in Chinese studies.In this paper,a human action recognition algorithm(AE-HRNet)based on high-resolution network(HRNet)and attention mechanism is proposed to address the problem of inadequate extraction of semantic and location information of human action features by convolutional neural networks.Firstly,the channel attention(ECA)module and spatial attention(ESA)module are introduced to increase the weight of local key point information in image features and reduce the loss caused by key point localization.In addition,a fusion output method is proposed to fuse the feature maps layer by layer from low to high levels,reducing the complexity of the operation while obtaining more adequate semantic and location information in the feature maps.Experimental results on the MPII and COCO validation sets in the same environment configuration show that AE-HRNet reduces the computational complexity and improves the accuracy of human action recognition compared to high-resolution networks.In this paper,we design a video-based human action recognition algorithm based on AE-HRNet to explore the application of human action recognition in the field of home education.Firstly,the video data is detected based on the target detection algorithm,and if the detection is successful,a separate screenshot is taken and tracked and processed into picture data.The AE-HRNet network is used for human action recognition,and the content of Chinese culture education corresponding to the recognition result is matched according to the rule-based intelligent push law and fed back to the user in the form of speech.The application can create an immersive Chinese phylosophy education atmosphere by automatically acquiring human movement information without violating the user’s privacy.