一种基于快速面部识别算法的学习状态统计分析系统  

A statistical analysis system of learning state based on fast expression recognition algorithms

作　　者：王珺[1] WANG Jun(Xi’an Aeronautical Polytechnic Institute,Xi’an 710089,China)

机构地区：[1]西安航空职业技术学院,陕西西安710089

出　　处：《电子设计工程》2020年第5期36-39,46,共5页Electronic Design Engineering

基　　金：陕西省教育科学规划课题(SGH17V013)。

摘　　要：针对计算机智能辅助教学的需求,文中设计了基于人脸表情的学习状态统计分析系统。文中首先对人脸表情识别方法进行研究,并基于深度学习算法设计了系统的算法模块。该模块使用FEYOLO网络结构,同时在池化层与卷积层间加入Inception结构,减少网络参数1 620个,大幅度降低了网络的复杂度;网络使用SELU函数作为传递函数,缓解了在网络训练过程中的梯度爆炸。算法引入学习状态分析系统后,在自建的学生学习状态数据集上,经过对比实验表明,对于学生学习状态判别的准确度大于93%,优于系统现在使用的YOLO和VGG16网络。Aiming at the demand of intelligent assisted instruction,this paper studies the relevant algorithms of expression recognition based on deep learning technology.Firstly,the method of facial expression recognition is studied,and the algorithm module of the system is designed based on the deep learning algorithm.The module adopts FEYOLO network structure and inception structure,which reduces 1620 network parameters and greatly reduces the complexity of the network.In addition,the SELU function is used as the transfer function to alleviate the gradient explosion in the network training process.Through comparative experiments,the accuracy of learning state discrimination is more than 93%,which is better than YOLO and VGG16 networks.

关 键 词：深度学习 表情识别 FEYOLO 植入 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]