Research On The Application Of Face Recognition Technology In Teaching Management Based On Deep Learning

This thesis mainly discusses how to use the convolutional neural network to cont inuously observe students' learning status through real-time video recording,so as to provide teachers with teaching management.The traditional face recognition locate facial features by geometric method,whic h is easy to be obscured and leads to poor recognition effect.In addition to facing the camera face on,traditional face recognition also requires clear presentation of facial f eatures.However,in the actual class,students' faces may be obscured for various rea sons,so that they cannot be accurately identified continuously.Therefore,this paper proposes a convolutional neural network to extract image features for students' faces recognition.However,it is difficult to identify all the students because of the single camera s hooting,and it is easy to identify the characteristics of the background and cause misj udgment.Therefore,this study used Faster R-CNN to capture the face image in the fr ame and trained it through face recognition model.In addition to enabling multiple st udents in a picture to identify themselves,it can also eliminate unnecessary backgrou nd.The accuracy rate of the face recognition model trained in this study as high as 99%,and the test loss value is 0.0278137.In addition to the identity recognition,this pa per uses the emotion recognition model to analyze their emotions in class and record the statistics,so as to understand the students' mood in class.Finally,parameter drive r identification is passed through WebSocket,and a set of online teaching manageme nt system is established with FLUX architecture.The mainly research of this paper is to design a face recognition model using Fa ster R-CNN network and process research data,including parameter setting,sample c ollection and sample processing.Through the establishment of face recognition model in the early stage,the cross-platform system construction is realized,and the training results are obtained.Finally,the results are presented in the form of web pages for te achers to conduct teaching management evaluation.