Design And Implementation Of Postgraduate Enrollment Management System Based On FPGA Face Authentication

With the explosive growth of the number of applicants for the postgraduate entrance examinations and the transparency of postgraduate entrance examination information,the need for fairness and justice in the postgraduate admissions process is becoming more and more urgent,and the research and recruitment work is becoming more and more cumbersome.In order to ease the postgraduate recruitment work,an open,transparent and efficient Web information platform is urgently needed between the admissions unit and the candidates to complete the mutual selection.In addition,in order to ensure the information security of candidates,the design of the authentication system of the Web information platform is particularly important.As the unique information of each person,the face is more secure than the traditional username and password.Based on this,this paper designs a face authentication graduate admissions management system based on FPGA.The main contents are as follows:(1)A Web system with complete functions from the release of preliminary test results to the examinee’s intended admission was designed and implemented.It adopted the architecture of separating the front and back ends,in which Vue.js,Express and Socket framework were used for the front end,the back end and the communication between the back end and FPGA,respectively.(2)A lightweight face recognition system based on neural network was designed and implemented.The system uses the Face Net network to train neural networks.Combined with the characteristics of the Face Net network,the effects of different trunk networks and different feature dimensions on the accuracy of face recognition.(3)The deployment of face recognition system on FPGA was realized.Due to the limitations of FPGA computing characteristics and on-chip resource conditions,the network parameters were first quantified,and then the neural network circuit was optimized through loop block,loop unrolling,and function pipeline design,and finally deployed on the FPGA development board.The experimental results showed that the accuracy of face recognition based on FPGA was the same as that of CPU or GPU,but the computing time and power consumption of the former were greatly reducedThrough the postgraduate admissions management system effectively solved the problem of information asymmetry between colleges and candidates,and improved the efficiency of postgraduate admissions.In addition,the FPGA-based face recognition authentication system ensured the security of system data,and had the characteristics of low power consumption,fast response speed,and high cost performance.