Implementation Of The Online Design Thinking Tool And Analysis Of Collaborative Design Process

With the rapid development of society and the wide application of the Internet in recent years,online courses have gradually appeared in everyone’s field of vision.Online courses have the advantages of low cost,no contact,and long distance.However,the accompanying assessment measures often have drawbacks,and it is easy to be attacked by facial prostheses.The live detection technology is the key technology to ensure the safety of the assessment stage of online courses and to ensure that online courses can be used in practice.This thesis first summarizes the face detection algorithms,compares the differences between traditional face detection algorithms and deep learning algorithms,and finally proposes a face detection algorithm based on CNN,and optimizes the algorithm framework in terms of interpolation algorithms.,which improves the speed of the face detection algorithm;in addition,after summarizing the methods of face fraud,we design a feature fusion in vivo detection algorithm based on the traditional living detection algorithm.The main work of this thesis has the following three points:(1)After analyzing the defects of the traditional face detection Adaboost algorithm,YOLO,which can detect faces in real time,is selected as the face detection algorithm model of this system,and for the structural shortcomings of YOLO,Efficient is selected to optimize it.,and added the attention module,and finally chose the more traditional face recognition model,the Facenet network model.(2)Based on the actual application scenarios of this system,in order to make a good deployment on the webpage,the lightweight shufflenet was finally selected as the network structure of the live detection,and the idea of combining with the traditional living test method.Before the improved Shufflenet network,before entering the Shufflenet network,the characteristics of several basic features were combined with the characteristics of attachment weights.In addition,the SHufflenet network was optimized.(3)In the end,this thesis designs an online liveness detection system for MOOC system based on the face detection and liveness detection algorithms and engineering experience studied earlier.The designed system has no major errors in the testing process,and in practical applications Satisfactory results were obtained.