线上教学中基于隐私保护的情感交互模型研究  

Research on Emotional Interaction Model for Online Teaching Based on Privacy Protection

作　　者：于博[1] 陈浩[2] 胡东辉[2] YU Bo;CHEN Hao;HU Dong-hui(Information Construction and Development Center,Hefei University of Technology,Hefei 230009,China;School of Computer Science and Information Engineering,Hefei University of Technology,Hefei 230009,China)

机构地区：[1]合肥工业大学信息化建设与发展中心,安徽合肥230009 [2]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《软件导刊》2023年第1期12-19,共8页Software Guide

基　　金：高等学校区块链技术创新应用计划项目(2020qkl32);安徽省示范实验实训中心—网络安全与攻防实训中心项目(2020sxzx41)。

摘　　要：线上视频教学目前得到了广泛应用,但还存在一些弊端。在线上教学过程中,教师与学生无法进行有效的情感交互,学生的面部信息也存在泄露风险。分别基于情感识别神经网络、人脸融合、三维动画模型绑定的理论与技术提出3种解决方案,从不同技术路径实现线上教学中师生的情感互动,并实现学生面部隐私保护。方案一将SepConv2D神经网络对学生的情感分析结果映射到对应情感的emoji表情;方案二利用styleGAN神经网络生成不同样式的生成脸,并应用三角剖分算法和泊松融合算法实现生成脸与学生真实脸的人脸融合;方案三利用人脸关键点检测和Blender 3D动画模型绑定技术,实现3D动画模型与学生真实脸绑定。对上述方案的技术特点进行分析,推理出3种方案不同的适用场景,通过实验与问卷调查结果验证3种方案情感交互的可行性以及隐私保护的安全性。Online teaching is widely adopted,while it suffers from some drawbacks.In online teaching,teachers and students are unable to conduct effective emotional interaction,and students’facial information may be also exposed.Based on the theories of emotion recognition’s neural network,facial blending and 3D animated model binding,propose three solutions to realize the emotional interaction in online teaching,while at the same time protecting the student’s facial privacy.In the first scheme,it maps the emotional analysis results of students to the corresponding emojis by SepConv2D neural network.In the second scheme,styleGAN is employed to generate another facial style to be blended with students’faces with technic of triangulation and Poisson blending.In the third scheme,using the technics of the face marks detection and Blender 3D animated model binding,it binds the students’real face with 3D animated models.Based on the technical analysis,propose the different application scenarios for three schemes.The results of experiments and questionnaires demonstrate the emotional interaction feasibility and privacy protection security of the three schemes.

关 键 词：线上教学 情感交互 隐私保护 神经网络 人脸融合 三维动画模型绑定 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]